FormalPara Key Summary Points

Why carry out this study?

Patients requiring cataract surgery can have meibomian gland dysfunction (MGD) without presenting with dry eye symptoms.

Failure to address ocular surface disease (OSD) preoperatively could lead to refractive misses, fluctuating vision, induced higher-order aberrations, bothersome ocular symptoms, and new or worse OSD symptoms postoperatively.

This is the first study to evaluate the effect of treating MGD with a vectored thermal pulsation system (LipiFlow) prior to cataract removal and implantation of a range-of-vision IOL.

What was learned from this study?

This study shows measurable positive impact of LipiFlow treatment on visual symptoms in patients implanted with range-of-vision IOLs.

The use of LipiFlow before cataract surgery could optimize patient expectations from range-of-vision IOLs.

This study validates the recommendations of recent clinical consensus guidelines to proactively diagnose and treat ocular surface disease, even in the absence of dry eye symptoms, in patients with a cataract prior rather than post cataract surgery.

Introduction

Meibomian gland dysfunction (MGD) is defined as chronic diffuse abnormality of the meibomian glands (MG) and characterized by terminal duct obstruction and/or changes in glandular secretion [1]. MGD reduces availability of meibum to the lid margin and tear film. Thus, MGD can alter the tear film, causing eye irritation, inflammation, dry eye, and ocular surface disease (OSD) [1]. MG function and a healthy lipid layer are vital for ocular surface health [2].

Aging is associated with alterations in MG secretions with or without ocular symptoms or OSD diagnosis [3,4,5]. Accordingly, asymptomatic dry eye disease and OSD occur in up to 60–80% of patients presenting for cataract surgery [6]. Gupta et al. reported that 57% of patients presenting for cataract surgery had no previous diagnosis of OSD, but 81% had at least one abnormal tear film test, suggesting that OSD is often overlooked/underdiagnosed in this population [6]. This is a concern because abnormal tear film and/or corneal surface can cause errors in presurgical keratometric and topographic measurements [7, 8]. This can lead to IOL calculation errors, refractive misses, residual ametropia, and ultimately dissatisfaction in postoperative outcomes [6, 8]. A survey of clinicians reported that more than 90% of respondents felt that mild-to-moderate dry eye affected post-cataract surgery satisfaction [8]. Thus, MGD and OSD of any severity can lead to suboptimal visual outcomes after cataract surgery.

The American Society of Cataract and Refractive Surgery (ASCRS) conducted a survey of its membership and learned of educational gaps related to managing OSD in patients undergoing cataract and refractive surgery. The ASCRS Cornea Clinical Committee published guidelines, and recommends identifying and treating OSD prior to cataract surgery regardless of the presence of symptoms [8]. The Committee acknowledges that addressing OSD preoperatively can be cumbersome; however, they state that “its importance cannot be underestimated.” The committee concluded that failure to address OSD preoperatively could lead to refractive misses, fluctuating vision, induced higher-order aberrations, and new or worse OSD symptoms postoperatively [8]. Further, the committee recommends that any case of OSD, whether visually significant or not, should be prophylactically treated to prevent postoperative worsening.

The commonly recommended treatment of MGD is a twice-daily regimen of warm compresses and lid hygiene/massage. However, elderly patients often have poor compliance with these at-home treatments, and effective manual gland expression can be difficult [8]. An alternate treatment is vectored thermal pulsation (LipiFlow™; Johnson & Johnson Vision, Irvine, CA, USA), which applies constant heat and a sequence of pressure pulsations to evacuate static oils from the upper and lower MGs simultaneously and to improve glandular flow. The temperature safely and effectively heats the gland contents without causing thermal injury [2, 9]. Prior clinical studies demonstrate safety, effectiveness, and clinical utility of treatment with the LipiFlow in patients with MGD and dry eye symptoms [10,11,12]. LipiFlow is more effective than the commonly recommended treatment and the effectiveness is rapid, only requiring a single 12-min in-office procedure [2, 12,13,14]. As opposed to manual gland expression, LipiFlow causes an awareness of pressure without causing pain [13].

According to the ASCRS consensus guidelines, patients implanted with range-of-vision IOLs (i.e., multifocal or extended depth of focus, EDOF) may be more susceptible to visual disturbances from poor tear film or other OSDs and preoperative optimization of the tear film is valuable [8, 15]. Hence, we hypothesized that preoperative treatment of MGD with a vectored thermal pulsation system would help promote better postsurgical meibomian gland function, and improve ocular health and visual quality. Thus, the purpose of this study was to evaluate whether LipiFlow treatment prior to cataract surgery with Tecnis™ Symfony™ EDOF IOL implantation in patients with mild-to-moderate MGD safely improves postoperative outcomes.

Methods

Study Design and Patients

This was a prospective, randomized, open-label, crossover multicenter study conducted at five study sites in the USA. Prior to cataract surgery, the test group underwent LipiFlow treatment while the control group did not. Three months postoperatively, both groups were evaluated for clinical outcomes; then the control group received LipiFlow treatment as the crossover group. The control group was then evaluated at 4 months after surgery (1 month post LipiFlow treatment) for clinical outcomes. The study protocol was reviewed and approved by Salus Institutional Review Board in accordance with the Declaration of Helsinki and the International Council for Harmonization Guideline for Good Clinical Practice. All patients provided written informed consent and HIPAA regulations were followed. The study was registered on www.clinicaltrials.gov (NCT03708367).

Included patients were adults at least 22 years of age with bilateral mild to moderate MGD (defined as a score of 15 or less on a scale of 0 to 45), none to moderate dry eye symptoms bilaterally [defined as total score 0 to 15 on the standard patient evaluation of eye dryness (SPEED) questionnaire], and scheduled for bilateral cataract surgery with Tecnis Symfony IOL implantation. Key exclusion criteria were irregular corneal astigmatism; pupil or zonular abnormalities; having a systemic disease that causes dry eye; unwillingness/inability to abstain from using systemic antihistamines, systemic medications known to cause dryness, or prescription medications to treat MGD or dry eye; having prior intraocular, oculoplastic, corneal, or refractive surgery procedure; having ocular trauma, ocular infection, recurrent/active ocular inflammation, or punctal plugs or occlusion; eyelid abnormalities that affect lid function; having moderate to severe (grade 2–4) allergic, vernal, or giant papillary conjunctivitis; and pregnant or breast feeding.

Study Procedures

Patients were randomized 1:1 to receive the LipiFlow treatment preoperatively (test group) or after the 3-month postoperative visit (control group). Treatment with the LipiFlow system was conducted per the LipiFlow System Instructions for Use [16]. Patients were instructed on how to perform blinking exercises for 1 month after LipiFlow treatment to facilitate the flow of oils from the MGs into the tear film. Patients were instructed not to use any MGD or dry eye prescription medications or other treatments during the study (except over-the-counter artificial tears, ocular lubricants, ointments, emollients, or omega-3 dietary supplements). Except for LipiFlow treatment, no other MGD or dry eye treatment was prescribed or administered to patients throughout the study. Two to four weeks prior to cataract surgery, the test group received preoperative LipiFlow treatment according to manufacturer’s instructions and as reported by others [13, 16]. All subjects in both arms had bilateral cataract surgery via standard small-incision phacoemulsification and implantation of the Tecnis Symphony Non-Toric or Toric Extended Range of Vision IOL (Models ZXR00, ZXT150, ZXT225, ZXT300, and ZXT375).

The test and control groups were examined preoperatively to establish a baseline, and study assessments were collected at 1 and 3 months after surgery. The control group was also examined 1 month following the crossover/postoperative LipiFlow treatment (= 4-month postoperative visit). The following were assessed: visual acuity, manifest refraction spherical equivalent (MRSE), contrast acuity, visual symptoms via the Patient-Reported Visual Symptom Questionnaire (PRVSQ), biomicroscopic slit-lamp findings, ocular/visual symptom assessment (non-directed, spontaneous), corneal surface staining via fluorescein dye, conjunctival staining via lissamine green dye, complications, and adverse events (AEs). MG function was measured on the basis of secretion characterization via an MG Evaluator diagnostic instrument (Johnson & Johnson Vision, Irvine, CA) as described in the manufacturer’s instructions [17] and slit-lamp. Gland function was scored 0 = no secretion to 3 = clear liquid oil secretion, and the sum of scores for 15 glands in the lower eyelid was used to calculate the total MG secretion score (higher score indicates better gland function). Meibomian glands yielding liquid secretion (MGYLS) is a measure of gland expressibility, and was calculated as the count of the number of 15 glands with grade 2 or 3 function (total score range 0–15, and the higher the number the better the MG function). Tear breakup time (TBUT) was measured as an average of three measures, with a higher number indicating better tear film stability. Eyelid margin was evaluated via slit lamp or digital images from LipiView™ II Ocular Surface Interferometer (Johnson & Johnson Vision, Irvine, CA) digital imaging. The frequency and severity of dry eye symptoms were assessed via the SPEED questionnaire, with frequency scored 0 = never to 3 = constant and severity scored 0 = no problems to 4 = intolerable. The maximum SPEED score was 28 points, and a lower total score indicated less frequent/less severe symptoms. Although the study was not masked, to maintain consistency, one individual at each site conducted all study-related vision testing.

Outcome Measures and Data Analysis

The co-primary effectiveness endpoints were (1) mean monocular uncorrected distance visual acuity (UCDVA) at 3 months after surgery; (2) rate of refractive predictability at 3 months after surgery; (3) rate of bothersome ocular symptoms (PRVSQ) at 3 months after surgery; and (4) the mean change in total MG score in the test group compared to the control group from baseline to 1 month after surgery. In addition, the mean change from baseline to 1 month after surgery was evaluated for MGYLS, ocular surface stain grade, TBUT, and eyelid margin evaluation. The mean change from baseline to 3 months after surgery was evaluated for SPEED score and total MG score. In the control group, the change from 3 to 4 months after surgery (i.e., before vs. after crossover) was evaluated for UCDVA, BCDVA, total MG score, and total SPEED score. Safety was assessed by monitoring AEs and medical/lens findings.

The sample size determination was based on the primary endpoint of UCDVA. There was 90% power to detect a 1-line or greater difference in mean UCDVA between the test and control groups with 55 subjects in each group. This assumed one-sided two-sample t test with an alpha of 0.05 and standard deviation of 1.6 lines. Considering a 20% screen failure/dropout rate, 69 subjects per group were planned for enrollment. No more than 30% of enrolled subjects with no dry eye symptoms were permitted at each site. Continuous and categorical variables were summarized with descriptive statistics and frequencies and percentages. The mean change in total MG score was compared between groups at 1 month after surgery. All other key study endpoints were evaluated at 3 months after surgery. Additional endpoints were evaluated at 1 and 3 months after surgery, and at 4 months after surgery for the crossover control group only. Missing data were not imputed. Statistically significant difference was recorded when P ≤ 0.05. Data were analyzed using SAS (v9.4, SAS Institute).

Results

A total of 121 subjects were randomized, with 117 eyes treated in the test group and 115 eyes treated in the control group. One subject in each group was lost to follow-up. The demographics were similar between groups (Table 1). The mean age of the total population was 65.2 ± 7.7 years, 59.0% (69/117) were women, and 76.9% (90/117) were White.

Table 1 Demographics
Full size table

Visual Outcomes

A comparison of the visual outcomes at 3 months is presented in Table 2. As expected, the mean monocular UCDVA at 3 months was not significantly different between groups (co-primary endpoint). At the 3-month postoperative visit, the test group had significantly better mean monocular BCDVA than the control group (P = 0.0495; − 0.03 ± 0.09 vs. − 0.05 ± 0.08 logMAR, respectively), while the mean binocular BCDVA was similar between groups. The rate of refractive predictability (i.e., within 0.50 D and 1.00 D of target refraction) at 3 months was similar between groups, and both groups had mean achieved MRSE of − 0.24 D at 3 months (co-primary endpoint). At the 3-month visit, contrast acuity was similar between groups.

Table 2 Visual outcomes at 3 months after surgery
Full size table

Ocular Symptoms

The rate of bothersome ocular symptoms is presented in Table 3 (co-primary endpoint). The test group had a significantly lower incidence of being bothered by halos compared with the control group (P = 0.019; 58.6% vs. 78.95%, respectively). The control group had a significantly lower incidence of being bothered by multiple or double vision compared with the test group (P = 0.016; 8.8% vs. 25.9%, respectively). There were no significant differences between groups in the following bothersome symptoms: starbursts, sensitivity to light, glare related to scattered light, and poor low-light vision.

Table 3 PRVSQ rating at 3 months after surgery
Full size table

Ocular Surface Assessment

The total MG score change from baseline to the 1-month postoperative visit was not significantly different between groups (co-primary outcome, Table 4). However, at 3 months after surgery, the test group had a significantly larger improvement from baseline in total MG score compared with the control group (P = 0.046; 7.3 ± 9.3 vs. 4.7 ± 10.1, respectively). The test group had an improvement in SPEED outcomes but was not statistically different from the control group.

Table 4 Ocular surface assessment
Full size table

Anterior Ocular Health

Table 5 summarizes the mean change from baseline to 1 month after surgery in anterior ocular health outcomes. At 1 month after surgery, the test group had a significant decrease in corneal and conjunctival staining compared to the control group (corneal: P = 0.04; − 0.57 ± 2.33 vs. 0.20 ± 3.19, respectively and conjunctival: P = 0.002; − 1.2 ± 3.8 vs. 0.45 ± 4.03, respectively). There were no significant between-group differences in the mean change in MGYLS, eyelid margin evaluation, and TBUT at 1 month after surgery.

Table 5 Anterior ocular health scores: change from baseline to 1 month after surgery
Full size table

Crossover Outcomes

The outcomes in the control group after postoperative LipiFlow treatment are summarized in Table 6. Monocular UCDVA significantly improved at the 4-month visit compared with the 3-month visit (P = 0.03; 0.05 ± 0.12 vs. 0.07 ± 0.13 logMAR, respectively), while monocular and binocular BCDVA remained unchanged. The total MG score was significantly improved at the 4-month visit compared with the 3-month visit (P < 0.0001, 16.1 ± 11.5 vs. 12.0 ± 10.6, respectively) and total SPEED score improved at the 4-month visit compared with the 3-month visit; however, the difference was not statistically significant.

Table 6 Outcomes in the control group after postoperative LipiFlow treatment (4-month visit)
Full size table

Safety

Ocular serious AEs (SAEs) occurred in 1/59 subjects (1.7%) in the test group (cystoid macular edema = 1 eye) and 4/58 subjects (6.9%) of the control group (cystoid macular edema = 3 eyes, anterior capsular phimosis = 1 eye). None of the ocular SAEs were related to the study devices. The test group had no AEs; 12.1% (7/58) of the control group had AEs and all were categorized as undesirable optical phenomena. Rates of anterior segment, anterior chamber, and posterior segment findings were similar between the groups.

Discussion

In patients with MGD, inflammation and other obvious signs of pathology may be absent; thus, MGD diagnosis may be overlooked [18]. Prophylactic treatment of MGD is recommended because there is an increased risk of worsening quality of MG secretions, decrease in meibum expressibility, and increase in dry eye symptoms after cataract surgery [6, 10, 19]. To our knowledge, this is the first study to evaluate visual outcomes and safety of treating MGD with a vectored thermal pulsation system prior to cataract removal and implantation of a range-of-vision IOL. Results of the co-primary endpoints show that LipiFlow treatment prior to cataract surgery with Symfony IOL implantation had no negative impact on 3-month postoperative objective visual outcome assessments. At 3 months after surgery, there was no difference in mean monocular UCDVA between study and control groups, and both groups achieved similar MRSE. In addition, the control group had no notable change in MRSE from 3 to 4 months after receiving postoperative LipiFlow treatment. However, there were notable differences in the rate of subjective bothersome ocular symptoms. Halos are the most frequently reported ocular symptom after presbyopia-correcting IOL implantation [20]. As anticipated, the test group had a significantly lower incidence of bothersome halos 3 months after surgery, which may be attributed to a healthier tear film post-LipiFlow treatment and less subsequent light scatter (i.e., fewer halos). In contrast, the control group reported a significantly lower incidence of multiple or double vision compared to the test group (Table 3). This finding could be attributed to the difficultly patients have in recognizing and understanding the difference between blur/shadowing of a letter and diplopia [21]. Shippman et al. reported that many patients who present with symptoms of double vision do not have diplopia [21]. Thus, it is possible that the patients had difficulty interpreting the PRVSQ multiple/double vision questions, and that the patients did not have true diplopia. Although 25.86% of the test group reported multiple/double vision over the last 7 days, PRVSQ Q3E data revealed that only 5.17% of patients reported that the multiple/double vision was causing difficulty. This response was not significantly different from the control group (P = 0.317). This suggests that the patients noticed the visual disturbance, but it did not affect day-to-day activities.

No safety concerns or relevant safety findings were uncovered. Everything was routine and as expected after cataract surgery. Both groups had similar surgeries and postsurgery treatment, so it would be expected that all eyes have some surgery-induced dry eye [11, 22]. The healing process varies between patients, and the healing process affects dry eye and MGD. One month after surgery may have been too soon to examine changes in the eyelid margin, total MG score, and MGYLS, thereby explaining the lack of significant difference between groups at this timepoint (Table 5). In contrast, at 3 months, the test group had a significantly larger improvement than the control group in total MG score (Table 4). The improvements from baseline in both SPEED scores and TBUT were greater in the test group than the control group; however, the between-group differences were not statistically significant. This may be attributed to either environmental factors [23], which are known to impact tear film and drive symptoms, or to the timepoint for evaluation. Park et al. reported that a significant between-group difference in TBUT following LipiFlow and cataract surgery with a monofocal IOL was not evident until 3 months postsurgery [19]. Despite the study limitation of evaluating the anterior ocular health only at 1 month, the test group had significantly less corneal and conjunctival staining than the control group (Table 5), which indicates that the test group had a better ocular surface environment.

A study evaluating the effect of LipiFlow treatment prior to cataract surgery with implantation of a monofocal IOL concluded that severity of MGD at baseline was correlated with greater improvement in expressibility (MGYLS) and quality of meibum at 3 months after surgery [19]. It is noteworthy that our study excluded patients with severe MGD, whereas the Park et al. study did not. We similarly demonstrated a significant improvement in meibum quality at 3 months after surgery in the test group, but the lack of significant difference in MGYLS in our study could be related to the assessment at 1 month or that our study population excluded patients with severe MGD.

The ASCRS Cornea Clinical Committee concluded that ocular surface health cannot be maintained or rehabilitated in the absence of healthy MG function because MGs are the foundation of a healthy tear film, stable vision, and ocular comfort [8]. Dissatisfaction with visual outcomes following multifocal lens implantation can be caused by dry eye [24]. The ASCRS Committee concluded that patients who are dissatisfied with their visual outcomes following a range-of-vision IOL implantation should be examined and treated for OSD before undergoing Nd:YAG capsulotomy or IOL exchange [8]. Our data supports this suggestion; patients in the crossover group who had MG treatment after cataract surgery had significant improvement in UCDVA and total MG score (Table 6).

Optimal functioning of range-of-vision IOLs requires best astigmatic correction because as little as 0.50 D of astigmatism could result in blurry vision, halos, and ghosting [25]. Unstable tear film affects the quality of optical surface reflections from the cornea, which compromises keratometry and can affect the accuracy of measurement of the magnitude and axis of astigmatism [25]. Matossian et al. evaluated LipiFlow treatment prior to cataract surgery with a monofocal IOL in patients with MGD-associated dry eye. The study concluded that stabilization of the tear film following LipiFlow treatment altered the magnitude of astigmatism, emphasizing the importance of managing dry eye prior to determining the surgical plan for astigmatism management. This is especially true when implanting a range-of-vision IOL.

This study has some limitations. The baseline characteristic of having visually significant OSD was not collected; thus, there was no subgroup analysis evaluating patients who had visually significant OSD versus visually non-significant OSD. Possibly, the differences between test and control groups may have been larger if subgrouped by visually significant OSD. One of the co-primary endpoints was an evaluation of bothersome ocular symptoms as assessed by the PRVSQ. However, PRVSQ was not designed to evaluate patients after cataract surgery. PRVSQ was self-administered by the patients to minimize any effect of the doctor–patient relationship. Some studies show that improvement in meibomian function following LipiFlow may not be seen until several months post-treatment and treatment may improve over time [11, 12, 19]. In the present study, 1 month was chosen as the primary timepoint because postoperative drops were discontinued at 1 month. Greater improvements in MGD outcomes may have occurred beyond the duration of assessment. Nonetheless, significantly less corneal and conjunctival staining in the test group at 1 month indicates a better ocular surface environment.

Despite the study limitations, the study has a robust study design and demonstrates the value of preoperative and postoperative LipiFlow treatment. A benefit of the study design is that not all patients had dry eye symptoms; therefore, the study included a mixed population (up to 30% could be asymptomatic), which is reminiscent of a real-world practice. Note that all patients were assessed as having MGD in both eyes as per the study protocol.

Conclusion

Presurgical LipiFlow treatment of patients implanted with range-of-vision IOLs improved meibomian gland function and improved postoperative anterior ocular health. In addition, LipiFlow treatment reduced postoperative reports of halos and had measurable improvements in visual acuity. These findings support the ASCRS guidelines recommending the proactive diagnosis and management of MGD in patients with cataracts by assessing all preoperative patients for MGD and proactively treating MGD prior to surgery to improve patient experience. The use of LipiFlow before cataract surgery could optimize patient satisfaction from a range-of-vision IOL by producing a healthier more stable tear film to better support range-of-vision IOL outcomes.