A multicenter prospective study (REF#7546-D0609175A), approved by Sterling Institutional Review Board) was performed from October 2019 to June 2022. There was no ethics committee approval. The inclusion criteria included; 1) age 18 -75; 2) Tear breakup time (TBUT) ≤10 seconds; 3) evidence of meibomian gland (MG) obstruction based on total MGS of ≤12 in lower eyelids for each eye; 4) subjective symptom score (using the Standard Patient Evaluation of Eye Dryness (SPEED) questionnaire) ≥10; 5) At least one meibomian gland opening with a visible plugging over the eyelid margin; 6) no ocular pathology requiring treatment other than eye lubricant and conventional eyelid hygiene within the last month and during the study; and 7) Ability to abstain from dry eye/MGD medications or any device treatments for the time between the treatment visit and the final study visit. Patients were excluded if they had any intraocular inflammation, previous ocular surgery, ocular trauma in the past 12 months, ocular infection, eyelid structural abnormality, systemic diseases that lead to dry eye disease, skin cancer, pregnancy, or lactation. Patients who wore contact lenses were not included in the study.

Patients were screened for the study using the SPEED questionnaire, OSDI, TBUT measurement, meibomian gland score (MGS), and corneal fluorescein staining (CFS) assessment. Patients who satisfied the selection criteria were informed about the study and provided written consent to participate. The ethical management of patients and data was performed in keeping with the Helsinki protocol.

The Envision platform equipped with the Forma-I handpiece (InMode Ltd., Yokneam, Israel) was used to provide transcutaneous radiofrequency (RF) energy The Forma-I handpiece was specifically designed to treat periorbital skin as shown in Fig. (1). The distance between the two electrodes limits RF penetration depth to 2.5 mm. Ultrasound gel was used to aid in the conduction of energy into the tissue. Using this technique, the tissue was not ablated nor an open wound created. The handpiece monitored the temperature and controlled the RF energy to maintain the heat at predetermined levels in the range of 40 - 43°C. After cleaning the treatment area, approximately 2-3 mm of clear ultrasound gel was applied. Treatment parameters were set according to the subject's tolerance (Energy 5-20). Cut-off temperatures were set between 41.5° - 42°C, starting at 40 degrees, and gradually increased according to the subject's tolerance. Full contact of the handpiece with the skin with a slight pressure was maintained, with a hand flattening and stretching the treatment area as needed. After pressing the footswitch to initiate the RF energy the handpiece was moved from tragus to tragus back and forth with gentle pressure over the full treated zone to reach uniform skin heating. The movement was constant to avoid hot spots. The operator then retracted the brow and forehead skin to allow exposure and treatment of the upper eyelids. The treatment started at a low setting level to monitor the patient’s tolerance to the treatment parameters. The total treatment time was 12 minutes per side after which, a drop of topical anesthetic was placed into both eyes and therapeutic meibomian gland expression was performed at the slit lamp with Collins Expressor Forceps (Bruder). At one study site a protective eye shield utilizing a flexible handle was used for further ocular protection and to serve as an additional base for gland expression. With the shield over the anesthetized eye, the tip of the handpiece was able to treat along the eyelid margin for heating and then used to compress the outer lid and “ski” down the tarsal plate to express the oils while the lid was at maximal heat from the RF device. This technique had been described using a monopolar RF device compared to Lipiflow (Pelleve, an Ellman device no longer in production) [3]. Standardized outcome measurements were collected at baseline, 1 month, 3 months, and 6 months thereafter.

In all, 47 patients were recruited and completed treatment with the Forma-I handpiece, but not all subjects had complete data on all observations. When an observation was missing, the entire patient's data for that specific outcome was not included in the analysis. The study was performed at three separate institutions by three independent surgeons (Department of Plastic Surgery, Lenox Hill Hospital, New York, New York, USA, Department of Oculoplastic Surgery, Austin Face and Body, Austin, Texas, USA, and Excellent Vision Dry Eye and Rejuvenation Center in Portsmouth, NH, USA). These sites participated in the research, with measurements taken at baseline, 1-month, 3-months, and 6-months follow-up visits. The average patient age was 47.79 years, with the minimum age being 23 and the maximum age being 75 years. Of the 43 patients with gender data, 74% were female and 26% were male. Table 1 provides a summary of all primary endpoint measurements.

Thirty-two patients had complete baseline, 1-month, 3-month, and 6-month follow-up SPEED data. SPEED was assessed via a scale anchored by 0-4 (mild symptoms) and 8+ (severe symptoms). The within-subjects ANOVA was statistically significant, F(3, 99) = 23.42, p<0.001. The relationship was linear and consulting the pairwise comparisons demonstrated where those differences between periods occurred. 1-month, 3-month, and 6-month follow-up values were statistically lower than baseline values (Fig. 2). The 1-month values were also statistically higher than the 3-month values. In other words, SPEED improved from baseline to 1 month, and from 1 month to 3 months, before leveling off and maintaining that improvement through the 6-month follow-up period.

SPEED responders were those who had a 4 point improvement from baseline to the 6-month. Of the 32 patients with the 6-month data, 56% were responders. Additionally, extreme responders were also identified. This group of patients had severe dry eye symptoms at baseline and mild dry eye symptoms at 6 months. Only a few patients met this threshold (N = 2, or 6%).

Fig. (2). Standard Patient Evaluation of Eye Dryness (SPEED) measurements at baseline, 1 month, 3 month, 6 months after treatment. There was a statistical improvement from baseline to 1 month, and from 1 month to 3months, before leveling off and maintaining that improvement through the 6 month follow up period, (p<0.001). SPEED was assessed via a scale anchored by 0-4 (mild symptoms) and 8+ (severe symptoms).

Fig. (3). Ocular Surface Disease Index (OSDI) measurements revealed an improvement from baseline to 1 month, and this was maintained through the 6 month follow up time period, (p<0.1). ODSI was assessed via a scale anchored by 0-12 (normal) and 33+ (severe disease).

In this category, 35 patients had complete baseline, 1-month, 3-month, and 6-month follow-up data. ODSI was assessed via a scale anchored by 0-12 (normal) and 33+ (severe disease). The within-subjects ANOVA was statistically significant, F(3, 99) = 7.84, p = 0.01. The relationship was linear, and consulting the pairwise comparisons demonstrated where those differences among periods occurred. Follow-up values of 1-month, 3-month, and 6-month were statistically lower than baseline values; all follow-up visit results were statistically flat to one another. In other words, ODSI improved from baseline to 1 month, and this improvement was maintained through the 6-month follow-up period (Fig. 3).

ODSI responders were defined as those who had a 10-point improvement from baseline to 6 months. Of the 35 patients with 6-month follow-up data, 51% were responders. Additionally, extreme responders were also identified. This group of patients had severe dry eye symptoms at baseline and mild or normal dry eye symptoms at 6 months. Seventeen patients had severe symptoms at baseline, and 29% of those patients improved to mild or normal symptoms at 6 months.

Thirty-five patients had complete baseline, 1-month, 3-month, and 6-month follow-up data. CFS was assessed via a scale anchored by 0-3 (normal) and 12-15 (epithelial defect or large coalescent erosion). The within-subjects ANOVA was statistically significant, F(3, 99) = 65.01, p<0.001. The relationship was linear, and consulting the pairwise comparisons demonstrated where those differences between periods occurred. Follow-up values of 1-month, 3-month, and 6-month were statistically lower than baseline values. Thus, the data showed that CFS right eye improved from baseline to 1-month post procedure this improvement was maintained at each follow-up visit (Fig. 4a).

CFS right eye responders were defined as those who had a 2-point improvement from baseline to 6 months. Of the thirty-five patients with 6-month data, 80% were responders. A subset of these patients was classified as extreme responders or those that had an improvement in at least one category from baseline to 6 months. Of the applicable 35 patients, 23% were extreme responders.

CFS metrics were similarly assessed for the left eye. Again, the within-subjects ANOVA was statistically significant, F(3, 99) = 36.26, p<0.001. The relationship was linear, and consulting the pairwise comparisons demonstrated where those differences between periods occurred. As seen with the right eye, for the left eye, 1-month, 3-month, and 6-month follow-up values were statistically lower than baseline values (Fig. 4b). Again, the data showed that CFS left eye improved from baseline to 1 month, and was maintained through 6 months of follow-up.

Of the thirty-one patients with CFS left eye 6-month data, 84% were considered responders. Additionally, extreme responders were also identified. An extreme responder is a patient that has an improvement of at least one category from baseline to 6 months. Of the applicable thirty-one patients, a total of 29% were extreme responders.

Fig. (4a and 4b). Corneal Fluorescence Score (CFS) measurements for right eye (a) and left eye (b) at baseline, 1 month, 3 month, 6 months after treatment. CFS showed a statistically significant improvement from baseline to 1 month, and this improvement was maintained through the 6 month visit in both eyes, (p<0.001). CFS was assessed via a scale anchored by 0-3 (normal) and 12-15 (epithelial defect).

The MGS for the right eye was assessed via an additive scale of 15 measurements per eye; the scale was anchored by 0 (no secretion) and 3 (clear liquid). Thirty-five patients had complete baseline, 1-month, 3-month, and 6-month follow-up data. The within-subjects ANOVA was statistically significant, F(3, 99) = 53.59, p<0.001. The relationship was linear, and consulting the pairwise comparisons demonstrated where those differences between periods occurred. Follow-up values of 1- month, 3-month, and 6-month were statistically higher than baseline values (Fig. 5a). The 3-month values were also statistically higher than baseline and 1-month values, indicating that the MGS for the right eye improved from baseline to 1 month, and from 1 month to 3 months before stabilizing.

MGS responders were defined as those who had a 5-point improvement from baseline to 6 months post-procedure for the right eye. Of the thirty-five patients with 6-month data, 94% were considered responders.

Similar findings were observed for MGS in the left eye (Fig. 5b). Again, thirty-five patients had complete data at baseline and all follow-up visits. The within-subjects ANOVA was statistically significant, F(3, 99) = 45.21, p<0.001. The relationship was linear, and consulting the pairwise comparisons demonstrated where those differences between periods occurred. The MGS values were statistically higher at the 1, 3, and 6-month follow-up visits when compared to baseline. The 3 and 6-month values were also statistically higher than baseline and 1-month values, thus indicating that the MGS improved from baseline to 1 month and then from 1 month to 3 months. It was then stable from 3 months to 6 months after the procedure.

Responders for the left eye were again defined as those who had a 5-point improvement from baseline to 6 months. Of the thirty-five patients with 6-month data, 91% were responders.

TBUT measurements were assessed independently in the right (Fig. 6a) and left (Fig. 6b) eyes. For each eye, thirty-five patients had complete baseline, 1-month, 3-month, and 6-month follow-up data. TBUT was assessed via a scale anchored by less than 5s (dry eye) and greater than or equal to 10s (normal). The within-subjects ANOVA was statistically significant for the right eye, F(3, 99) = 20.63, p<0.001, and for the left eye F(3, 99) = 21.54, p<0.001. The relationship was linear in each and consulting the pairwise comparisons demonstrated where those differences between periods occurred. The data showed a statistically significant increase in TBUT at 1, 3, and 6 months after the procedure for both eyes. The TBUT improved from baseline to the 1-month visit and was still stable at the 3 and 6-month visits for both eyes.

TBUT responders were those who had at least a 3-second improvement from baseline to 6 months. Of the thirty-five patients with 6-month data, 60% were responders in the right eye and 69% were considered responders in the left eye. Extreme responders were identified as patients that had a TBUT of 5 seconds or less at baseline and a TBUT of 10 seconds or more at 6 months. Of the applicable 35 patients, 20% were extreme responders in the right eye and 17% in the left eye.

Fig. (5a and 5b). Meibomian Gland Score (MGS) for the right eye (a) and left eye (b) are shown at baseline, 1 month, 3 month, 6 months after treatment. MGS was assessed via an additive scale of 15 measurements per eye, and the scale was anchored by 0 (no secretion) and 3 (clear liquid). In both eyes, the MGS improved statistically at the 1 month follow up visit and this improvement was maintained out to the 6 month visit, (p<0.001).

Fig. (6a and 6b). Tear Break Up Time (TBUT) was assessed via a scale anchored by less than 5 (dry eye) and greater than or equal to 10 (normal) in the right (a) and left (b) eyes. A statistically significant improvement was noted at 1 month compared to baseline, and this improvement carried through to the 3 and 6 month follow up visits, (p<0.001)

Participants were asked to assess their improvement after 1, 3, and 6 months of treatment, anchored by 0 (no difference) and 4 (significantly marked improvement). The within-subjects ANOVA was statistically significant, F(2, 66) = 6.18, p = 0.01. Consulting the pairwise comparisons demonstrated that at 6 months, patients were more likely to believe they had improved as compared to 1 month and 3 months (Fig. 7). Additionally, all three improvement measurements were compared to the bottom value of the scale, 0, which indicates no improvement. All three median values were statistically higher than 0, which indicated that at each period, participants believed they saw some level of improvement. Taken together, participants were sufficiently satisfied with their improvement at 1-month, 3 months, and 6 months compared to believing there was no difference.

Participants were asked to assess their satisfaction after 1, 3, and 6 months of treatment, anchored by -2 (very dissatisfied) and 2 (very satisfied). The within-subjects ANOVA was not statistically significant, F(2, 66) = 0.85, p = 0.43. Additionally, all three improvement measurements were compared to the middle value of the scale, 0, which indicates neutral satisfaction. All three mean values were statistically higher than 0, which indicated that at each period, participants believed they were satisfied beyond a neutral level (Fig. 8). Taken together, participants were sufficiently satisfied with their improvement at 1 month and maintained that perception of improvement at 3, and 6 months.

For a follow-up period of 6 months, there were no skin burns or corneal trauma. Patients noted mild erythema of the periorbital skin that resolved in 24-48 hours. There were no complaints of a burning sensation during the procedure. No treatment was abandoned as no intolerance or excessive thermal sensation was reported.