Ocular surface health is dictated by a stable tear film. Instability of the preocular tear film is the hallmark of dry eye syndrome.¹ Ample evidence indicates that the stability of the preocular tear film in the interpalpebral surface (the first compartment) requires tears to be spread from superior and inferior tear menisci (the second compartment) through blinking.^{2, 3, 4 and 5} Little is appreciated if tears in the conjunctival sac (the third compartment) also contribute to the stability of the preocular tear film. By applying fluorescein to the superior bulbar conjunctiva under the lid and measuring fluorescein clearance in the meniscus, Mishima et al in 1966⁶ determined an average of total tear volume to be 6.2 μl, of which one half is estimated in the first 2 compartments and the other half in the third compartment. By applying TC^99m to the tarsal conjunctiva, Fraunfelder in 1976⁷ demonstrated rapid appearance of the tracer in the nearest meniscus. These 2 studies clearly indicated the presence of a tear flow from the third to the second compartment. By applying fluorescein to the base of the inferior fornix, MacDonald and Maurice in 1991⁸concluded that such a tear flow from the fornix to the meniscus is not mediated by diffusion. Without providing any explanation, all 3 studies revealed slower fluorescein clearance from the meniscus,⁶ longer entrapment of TC^99m in the conjunctival sac,⁷ and slower appearance of fluorescein in the precorneal tear film⁸ in older study subjects.

Conjunctivochalasis (CCh) is an age-dependent, bilateral eye disease characterized by redundant and loose conjunctival folds interspersed between the globe and eyelids.⁹ Previous studies have shown that conjunctival folds in CCh may cause delayed tear clearance by occluding the punctum^{10, 11 and 12} and interrupting the tear meniscus.^13 and 14 Nonetheless, it remains unclear whether conjunctival folds in CCh may also interfere with the tear flow from the fornix to the meniscus. Eyelid blinking, which plays a vital role in the spread and clearance of aqueous tears, also facilitates the tear flow from the fornix to the meniscus⁷and consequently the preocular tear film.⁸ However, for reasons that remain elusive, blinking, which normally helps relieve discomfort in patients suffering from aqueous tear deficiency (ATD) dry eye, does not help CCh patients.¹³ Furthermore, it is also puzzling that some patients with CCh are symptomatic whereas others are not.

To address these questions, we hypothesize that CCh also interferes with the tear flow from the fornix (the third compartment) to the first 2 compartments. Because CCh is prevalent in older people, this hypothesis, if proven, can help to explain why there is great variability in results obtained by the aforementioned 3 studies. Furthermore, we also hypothesize that symptoms ensue in patients with CCh if blinking is no longer sufficient to facilitate tear flow from the fornix to the tear meniscus. Herein, we have developed a new test to confirm these 2 hypotheses. Our findings provide a mechanistic understanding regarding the symptomatology of patients suffering from CCh and direct us to devise proper surgical correction.

This study was approved by the ethics committee of the Ocular Surface Research and Education Foundation (Miami, FL) and conducted after obtaining informed consent from all subjects and patients according to the Tenets of the Declaration of Helsinki. At the Ocular Surface Center (Miami, FL), we prospectively and consecutively enrolled a total of 24 CCh patients, who were seen between January and May 2012. We excluded CCh patients who had a past history of ocular trauma or who underwent eye operations other than for CCh. We also excluded patients with active ocular infection, allergy, punctal plugs, abnormal blinking, or ATD dry eye. Furthermore, we ensured that enrolled CCh patients did not use any topical medication or contact lenses for ≥1 week at the time of testing except those after CCh surgery. Among the 24 patients with CCh, 8 were asymptomatic and 16 had symptoms of ocular irritation and dryness. By the time of enrollment, 9 of 24 CCh patients had undergone operative correction by a single surgeon (SCGT) using a previously reported operative technique including excision of mobile and degenerated Tenon’s and fornix reconstruction with cryopreserved amniotic membrane (Bio-Tissue, Miami, FL) under topical anesthesia.^15 and 16 (Table 1, available at http://aaojournal.org). These 15 CCh patients without surgery were further divided into 2 subgroups: CCh without symptoms (54.5±10.4 years old; range, 38–66 years; n = 8) and CCh with symptoms (68.9±11.2 years old; range, 58–91; n = 7; Table 1). For comparison, we also enrolled as controls 13 normal subjects without ocular symptoms and diseases, who were subdivided into 2 subgroups: Young (30.5±4.4 years old; range, 25–38; n = 6) and old (58.9±16.4 years old; range, 44–85; n = 7; Table 1).

All subjects and patients had a routine history taken, including symptoms of eye irritation as well as both external and slit-lamp eye examinations without the use of any anesthetics and dyes. Specific attention was given to exclude abnormalities in cornea, conjunctiva, lids, and puncta. The diagnosis of CCh was based on the criteria previously summarized.⁹ After confirming that they met the inclusion and exclusion criteria, we carried out a special tear test (detailed below), followed by fluorescein staining and the fluorescein clearance test (FCT), of which the latter helps diagnose ATD with or without reflex tearing as well as delayed tear clearance.¹⁷

To test our hypotheses, we developed a special method to deplete the inferior tear meniscus and to monitor depletion and recovery rates by videorecording. After pulling down the lower lid, 5 μl of 0.25% fluorescein sodium with 0.4% benoxinate hydrochloride (Akorn, Buffalo Grove, IL) was applied to the base of the inferior fornix by an Eppendorf micropipette (Rainin Instrument, Oakland, CA) without touching any part of the eye (Fig 1A, available at http://aaojournal.org). After normal blinking to achieve maximal mixing of fluorescein, the tear meniscus height was visualized under a blue light projected at a 60° angle to the observation binoculars, which was set at 16× magnification of a slit-lamp biomicroscope (Fig 1B). Under videorecording, a vitreous capillary tube with 0.3-mm inside diameter (Charles Supper, Natick, MA) was used to deplete the inferior tear meniscus with a horizontal motion starting from the central portion while the subject looked straight ahead without blinking for as long as possible (Fig 1C). These steps were repeated 3 times without blinking. Then, the same test was repeated another 3 times immediately after a conscientious blink.

The recorded video displaying depletion and subsequent recovery of the inferior tear meniscus was analyzed by a media player (StormCodec Version: 5.09.0118.2111, Beijing, China), which continuously captured about 28 pictures every second during the playback. From the sequential captured pictures, we designated the time as zero when the tear meniscus was maximally depleted by the capillary tube. Thereafter, the length of the depleted inferior tear meniscus was measured. For videos without and with blinking, we monitored the recovery by analyzing captured pictures for 8 (Fig 2A) and 1.5 seconds (Fig 2B), respectively. Because blinks with abnormal duration might interfere tear kinetics, we excluded videos that exhibited blinking of which the duration exceeded 0.37 second or was <0.26 second, a normal range of voluntary blinking reported by Sun and Porter.¹⁸ To measure the tear meniscus height more precisely, we used Microsoft Office Picture Manager (Version, 12.0.6606.1000; Microsoft, Inc, Redmond, WA) to enlarge each picture 8 times of the original size, and measure the tear meniscus height at the 3 points projected from the center of cornea across 5, 6, and 7 o’clock, respectively (Fig 1D). The average of these 3 locations was used to designate as the tear meniscus height of that given time point. Using the tear meniscus height before depletion as the baseline, we calculated the rate of depletion and recovery as percentage for each time point.

Results were expressed as mean values ± standard deviation. For the study, only the left eye in normal subjects and the worse or operated eye in CCh patients were chosen (Table 1). The percentage of recovery of the original height of the inferior tear meniscus at each time point was compared among different subgroups, namely, normal subjects, asymptomatic CCh patients, and symptomatic CCh patients with or without surgery by analysis of variance. For multiple comparisons, we used the least significant difference procedure in analysis of variance. The comparison between senior and junior normal subjects and the difference of recovery rate between normal and CCh patients without surgery was performed by independent-samples t test. All these statistical analyses were performed using SPSS software version 11.5 (SPSS, Inc, Chicago, IL), and reported as 2-tailed probabilities. P<0.05 was considered significant.

As shown in Table 1, there were 7 males and 6 females of 45.7±19.1 years old (range, 25–85) in 13 normal subjects. The FCT result confirmed their tear function to be normal without delayed tear clearance. The use of a capillary tube along the tear meniscus depleted 98.7±1.5% of the width of and 39.8±2.0% of the original height. The meniscus height was rapidly recovered to 72.9±8.2 % and 71.6±4.3% in 3 seconds between young and old subgroups, respectively, without blinking (P>0.05; Table 2; Fig 3A), but was recovered to 72.1±4.6% and 73.1±3.0% in 0.6 second for both subgroups after 1 blink (P>0.05; Table 3; Fig 3B). These results indicated that blinking expedited the recovery of the tear meniscus height 5 times faster than that without blinking, and that such rapid recovery was not affected by aging in normal subjects.

In 15 CCh patients without surgery, 5 males and 10 females of 61.2±12.8 years old (range, 38–91), the capillary tube also depleted an average width of 66.5±14.8% of the entire inferior meniscus. Because of the preexisting nasal and temporal wrinkled conjunctiva, such maneuver maximally depleted the inferior tear meniscus and the tear meniscus height to 40.9±3.5% of the original height, which was not different from that of normal subjects (P>0.05). Compared with 13 normal subjects as a group, these 15 CCh patients showed a significantly lower recovery rate without blinking ( Table 2; all P = 0.000; from 1 to 8 seconds) and with blinking ( Table 3; all P<0.05; from 0.4 to 1.5 seconds). These results indicated that CCh patients as a whole exhibited a significantly slower recovery of the tear meniscus height after maximal depletion.

The symptoms of CCh patients resembled those of dry eye patients and included dryness, pain, foreign body sensation, burning, tearing, light sensitivity, and blurry vision. FCT showed that they all did not have ATD dry eye (Table 1). Without blinking, the speed of recovery in both subgroups was significantly slower than that of normal subjects (Table 2; all P<0.05; Fig 4A). The overall recovery rate of 7 CCh patients with symptoms was significantly less than that of 8 CCh patients without symptoms ( Table 2; all P<0.05; from 2 to 8 seconds). At the 8th second, CCh patients without symptoms reached 65.6±6.4%, which was significantly higher than 53.7±5.5% for CCh patients with symptoms ( Table 2; Fig 4A; P = 0.000; Video 1, available at http://aaojournal.org). After 1 blink, CCh patients without symptoms rapidly recovered the meniscus height to the same extent as normal subjects at each time point ( Table 3; all P>0.05; from 0 to 1.5 seconds; Fig 4B). In contrast, CCh patients with symptoms could not do so, and their recovery rates remained significantly slower than those of normal subjects and CCh patients without symptoms ( Table 3; all P<0.05; from 0.2 to 1.5 seconds; Fig 4B; Video 2, available at http://aaojournal.org).

Before surgery, another 9 CCh patients, 4 males and 5 females of 66.1±12.5 years old (range, 38–80), all had similar symptoms to those without receiving surgery (Table 1). They underwent surgical correction of CCh by excision of degenerated and mobile Tenon’s and reconstruction with amniotic membrane transplantation. Symptoms disappeared in 7 patients; the remaining 2 noted mild foreign body sensation on postoperative day 1, when the inferior fornix was deepened. The width of the tear meniscus was similarly depleted to 96.5±3.4%, which is not different from normal subjects (P>0.05). They all rapidly recovered the tear meniscus height to the same extent as normal subjects without blinking (all P>0.05; from 0 to 8 seconds; Table 2; Fig 4A) or with 1 blink (all P>0.05; from 0 to 0.8 seconds and 1.5th second;Table 3; Fig 4B). These results suggested a high correlation between symptomatic relief and rapid recovery of the tear meniscus height after maximal depletion in CCh patients as early as the first postoperative day when wound healing had not begun.

Previous studies using fluorescein^6 and 8 or TC^(99m)7 to trace the tear flow strongly support the presence of a tear flow from the fornix (the third compartment) to the tear meniscus (the second compartment)^6 and 7 and the preocular tear film (the first compartment).⁸ Using a capillary tube, we could not deplete >60% of the original height of the tear meniscus (Table 2; Fig 3), presumably because of a negative hydrostatic effect generated by the meniscus surface tension.^3 and 19 In normal subjects, recovery was rapid without blinking and achieved about 72% of the original height in <3 seconds. Such rapid recovery is unlikely caused by tear secretion at a rate of 0.02 μl per second, which can only replenish 4% of the entire inferior meniscus in 3 seconds, assuming that the tear volume of the inferior tear meniscus is about 1.45 μl.⁶ The ultimate recovery of the original tear meniscus depends on continuous secretion of aqueous tears, which takes a much longer time. Furthermore, such rapid recovery is also unlikely caused by reflex tearing because we used topical anesthetics in the fluorescein solution and avoided direct eye contact. Yokoi et al²⁰ observed rapid reduction and restoration of the meniscus radius after inserting a cotton thread or Schirmer’s strip in normal subjects and ATD dry eye patients. Because both the cotton thread and paper strip reached deep in the fornix and because the remaining tear meniscus was not depleted, the likely replenishing source was the adjacent meniscus. Because a capillary tube was used to deplete a much wider length of the tear meniscus and did not disturb the fornix, the replenishing source is more likely from the fornix. This interpretation is consistent with rapid appearance of TC^99m in the nearest tear meniscus when applied to the tarsal conjunctiva.⁷ Because recovery of the depleted meniscus was not different in those >44 years old when compared with those <38 years old (Table 2 and Table 3; Fig 3), aging is not a factor affecting such tear flow in normal subjects without CCh.

In contrast, recovery was significantly retarded in CCh patients (P<0.001) with or without symptoms, although a similar (e.g., 60%) maximal depletion of the original height was also achieved. The capillary tube surely depleted the entire inferior meniscus in CCh patients because of preexisting conjunctival folds. Thus, under such a circumstance, the replenishing source could come from either the preocular tear film (the first compartment) or the fornix (the third compartment). Because no recovery of the tear meniscus was noted in CCh patients for ≥8 seconds without blinking, we concluded that the preocular tear film cannot replenish the tear meniscus in CCh patients and most plausibly also in normal subjects. Because both normal subjects and CCh patients did not have ATD dry eye defined by FCT ( Table 1), we also concluded that the fornix, but not the preocular tear film, was the primary source of replenishing depleted tear meniscus after maximal depletion. Because conjunctival folds are detected in 90.2% of patients with aged 41 to 50 years and higher in a hospital-based study ²¹ and because normal subjects as old as 89 ⁶ and 67 years old ⁸ were included in their studies, our findings explain why there is high variability in the first phase of fluorescein clearance in the inferior tear meniscus ⁶ and in the duration to achieve the first appearance of fluorescence on the cornea. ⁸ Furthermore, it also explains why subjects >50 years old had a slower tear flow of TC^99m from the tarsal or fornix conjunctiva to the tear meniscus and a longer retention time in the fornix than those 20 to 30 years old. ⁷ Further studies are needed to determine which area of the fornix has the greatest capacity of holding tears as a way of explaining the severity of CCh.

Blinking plays an important role in tear spread and clearance,^22 and 23 thereby contributing greatly to tear film stability and elimination of potentially harmful cytokines, toxins, allergens, and microbes.^{12, 17, 24 and 25}The finding that blinking also increases tear flow from the fornix to the meniscus and ultimately to the preocular tear film has been reported by the aforementioned 3 studies.^{6, 7 and 8} Herein, we further showed that blinking facilitated recovery of the meniscus height after maximal depletion. In normal subjects, 1 blink recovered an average of about 72% of the original height in as little as 0.6 second, which was 5 times faster than that without blinking (Table 3; Fig 3). Symptoms of CCh patients can be caused by instable tear film, delayed tear clearance, and exposure.^{9, 10, 11, 12 and 13} Although the speed of recovery was significantly more retarded than normal subjects without blinking, 1 blink was sufficient to recover the tear meniscus height in asymptomatic CCh patients as rapidly as normal subjects, but not symptomatic CCh patients (Table 3; Fig 4). This finding suggests that blinking is an effective compensatory mechanism to restore the tear meniscus so as to keep some CCh patients asymptomatic. It also explains why increased blinking that surely stabilizes the tear film in ATD dry eye by shortening the interblink interval fails to help symptomatic CCh patients.¹³ If we assume that all CCh patients had the same extent of interference of the tear flow from the fornix, 1 important clinical parameter to distinguish them into symptomatic and asymptomatic is deficiency in blinking that is commonly noted in the older population.^{18, 26 and 27} Nonetheless, various extents of conjunctival folds to interfere with tear flow from the fornix may well exist in CCh. Conjunctival folds extending down to the fornix escape detection under routine slit-lamp examination because gravitation unwrinkles these folds when the subject is in an upright position. Loose conjunctival folds were best examined when the subject is in a supine position, for example, during the operation, which partially neutralizes gravitation.⁹ Our recent studies have revealed that conjunctival folds result from degeneration and dissolution of the Tenon’s capsule owing to apoptosis and overexpression of matrix metalloproteinase-1 and -3, as well as perturbation in expression of 2 anti-inflammatory genes—tumor necrosis factor-stimulated gene-6 and pentraxin 3—by conjunctival and Tenon’s fibroblasts.^28 and 29 Further studies are needed to determine how blinking actually facilitates the tear flow from the fornix to the meniscus.

Conjunctival folds in CCh could conceivably occupy and deplete the tear reservoir in the fornix (the third compartment), leading to a marked reduction of the overall tear volume/capacity. That explains why some symptomatic CCh patients do not benefit from frequent use of artificial tears and punctal occlusion and frequently complain of “tearing” (Table 1). Because punctal occlusion creates delayed tear clearance that might already be present in eyes with CCh,^{9, 10, 11, 12 and 13} it is generally considered contraindicated unless ocular surface inflammation is controlled. Taken together, the new mechanistic understanding of CCh unraveled herein also shed new light on how we should devise the surgical correction. Ideally, an effective operative procedure should eliminate conjunctival folds not only at the tear meniscus but also in the fornix. This notion is supported by rapid recovery of the original tear meniscus height with or without blinking in symptomatic CCh patients when they underwent excision of mobile and degenerated Tenon’s and fornix reconstruction by transplantation of cryopreserved amniotic membrane (Table 2 and Table 3; Fig 4). As noted, this explains why several studies have reported marked relief of symptoms by such operative procedures in CCh patients.^{15, 16, 30 and 31} That is also why surgical correction of CCh is advocated first to deepen the tear reservoir/volume and smooth out the tear meniscus before punctal occlusion for those with concomitant ATD dry eye.³² By the same token, one may imagine why obliteration of the fornix by symblepharon or other pathologies can present as a threat to the ocular surface by depleting the tear reservoir in the fornix, which in turn can destabilize the tear film. Deepening the fornix by symblepharon lysis and reconstruction is also an important measure to manage dry eye in a number of sight-threatening ocular surface cicatricial diseases.