A Systematic Review of the Efficacy of Microfocused Ultrasound for Facial Skin Tightening

Objective: to systematically review the efficacy of microfocused ultrasound (MFU) for facial skin tightening. Methods: A systematic search was performed (Pubmed, Embase) to assess the efficacy of single MFU treatments for facial skin tightening. Eligible studies included randomised controlled trials, controlled trials, cohort studies and case series (n ≥ 10). Objective and subjective outcomes were assessed. Results: A total of 693 studies were identified of which 16 studies were eligible. All the studies involved female patients. MFU is capable of tightening the skin, as observed in studies measuring the results of brow lifts (0.47–1.7 mm) and submental lifts (measured as a 26–45 mm2 reduction in the submental area on lateral photographs). Data from the Global Aesthetic Improvement Scale (GAIS) were pooled, and the day 90 pooled subjective investigator reported scores (IGAIS) (n = 337) showed that 92% of the patients demonstrated an improvement in skin tightening and/or in wrinkle reduction which continued up to one year. Longer-term follow-up data are not available. The patient-reported pooled scores (SGAIS) (n = 81) showed that the skin improvements were mild and continued to increase from 42% (90 days) to 53% (360 days) post-treatment. The MFU treatment was moderately painful and caused transient erythema with or without oedema. Other adverse effects were rare (2%), including dysesthesia (numbness or hypersensitivity), bruising and stinging, mandibular burns, striations and contact dermatitis. Various device settings, treatment protocols and energies were applied. Excessive skin laxity and a BMI > 30 were posed as relative contraindications for MFU treatment because positive results declined with an increase in laxity and BMI. Conclusions: MFU treatment is effective in tightening female patients’ mildly to moderately lax facial skin. Future studies should focus on objective treatment outcomes, optimising treatment regimens and male patients.


Introduction
The aim of microfocused ultrasound (MFU), a noninvasive treatment method, is to tighten the skin. It is an energy-based modality that induces tissue damage followed by tissue necrosis, whereby the energy of the ultrasound waves is converted into heat and cavitation [1]. The target areas are the subdermal connective tissues, such as the superficial muscular aponeurotic system (SMAS) layer and (deep) dermal layers. Multiple, small thermal injury zones (TIZ) of about 1 mm 3 in size are created at predetermined depths, the aim being not to damage the surrounding tissues [2,3]. TIZ result in immediate collagen contraction and denaturation that induce neocollagenesis and neoelastogenesis for more than one year [4][5][6]. Together, these processes are thought to contribute to the tightening of the skin for rejuvenation purposes. Although several studies in the literature measured the effect of MFU objectively and subjectively on skin tightening, more rigorous data are missing. Therefore, the literature on the efficacy of MFU for facial skin tightening was reviewed systematically.

Methods
This systematic review was prepared according to the Preferred Reporting Items for Systematic and Meta-Analysis (PRISMA) guidelines [7]. A search of the international prospective register of systematic reviews (Prospero) did not unearth any systematic reviews regarding this topic on the date of the search (19-08-2021). The protocol ID is 179974.

Search Strategy
The Pubmed and Embase databases were searched with these keywords: "high intensity focused ultrasound", "focused ultrasound" and the abbreviations "HIFU" (High-Intensity Focused Ultrasound), "IFUS" (Intense Focused Ultrasound), "MFU" as well as "skin", "tightening", "rejuvenation", "laxity", "cosmetic", "rhytids", or, "wrinkle", (Appendix A). The search was conducted to include the scope of ultrasound AND skin to select publications that deal with ultrasound and skin; to select publications that deal with skin rejuvenation more specifically, different words addressing skin rejuvenation were selected so that at least one of them would be present (OR). The reference lists of the included studies were screened to find any publications missing from the search.

Eligibility Criteria
Eligible studies were randomised controlled trials, controlled trials, cohort studies and case-control studies with ≥10 participants. The length of the follow-up had to be ≥3 months. No language restrictions were applied (Appendix B). The aim of the studies was an investigation of the efficacy of MFU on skin tightening, skin laxity and wrinkles after a single session of treatment. If the same treatment protocol was studied on different areas of the body but the face was evaluated separately, the study was also considered eligible for inclusion. When treatment protocols were combined in the same area, the study was excluded. The primary outcome measurement was skin improvement (measured as grades in improvement, skin tightening, skin laxity and/or wrinkles). The secondary outcome was adverse effects.

Study Selection
MC performed the search strategy and removed all duplicates. The selection of studies for inclusion was carried out by two observers (MC, MH) by first checking if any of the eligibility criteria appeared in the title or abstract. Then, the full texts of the selected articles were evaluated further. If there was uncertainty because the abstract or title did not provide sufficient information as to whether the selection criteria had been met, the full text articles were also evaluated. An insight into the selection decisions is depicted in a flow chart (Appendix C).

Data Extraction
The data were extracted from the included papers by one observer (MC) and noted on standardised extraction forms. The second observer (MH) carried out an independent check of the data extraction forms to ensure that all was complete and correct.

Data Synthesis
All the methods used to assess the effect of MFU with the selected studies were evaluated (for example brow lifting). The studies' results were pooled when possible. Descriptive statistics were used.

Study Selection
A total of 693 studies were identified for inclusion and screened after removing duplicates. There was a disagreement about including 23 studies based on the abstract. These studies were discussed in a consensus meeting and, if a disagreement persisted, a third observer (JS) was available to give a binding verdict. After the abstract screening, a total of 43 full texts were assessed. The excluded studies did not have any of the eligibility criteria that appeared in the title or abstract. The initial search was in June 2019, and the last update was in June 2021. In this period, another five studies were included, leading eventually to 16 studies being eligible for analysis (see the flowchart in Appendix C for the selection process).

Interobserver Agreement
After assessing the titles and abstracts, the agreement between the two observers (MC and MH) was 96%. A further discussion resulted in a consensus of 100% with a Cohen's kappa of 1.0.

Study Characteristics
The included study characteristics are displayed in Tables 1 and 2. All of the 16 included studies were eligible for qualitative analysis and six of them also for quantitative analysis (Table 3). Six studies evaluated the face as a whole whereas 12 studies analysed parts of the face (also) separately.       (Table 3) The objective measurements applied in the studies were diverse. Skin tightening after MFU was expressed as the number of millimetres of vertical displacement of the brows after treating the forehead and periorbital region [12,14,17,18,22], the number of square millimetres of submental lift after a full-face treatment [10,12,14] and the number of millimetres of oblique displacement of the marionette lines [18]. These measurements were done on 2D light photographs taken at baseline and different follow-up periods, typically 90 days (3 months), 180 days (6 months) and 360 days (1 year).
The absolute brow lifts varied from 0.47 mm (p < 0.02) [14] to 1.25 mm (p < 0.00) [22] and 1.7 mm at 3 months [17] post-treatment. Only one study gave 6-month results (1.22 mm) [22]. Sasaki et al. [18] depicted the distance between the baseline and brow edges as a percentage and reported a brow lift of 5.6-7.2% 6 months after applying the MFU treatment lines in a superolateral direction and a brow lift of 1-3.7% after applying the treatment lines in a horizontal direction. Werschler and Werschler [12] reported a percentage of patients with a brow lift of >0.5 mm. At 3 months, 35% of the subjects had a brow lift of more than 0.5 mm, and this had increased to 45% after 6 months but then decreased to 39% by the 1-year follow-up.
When treating the submental area, skin tightening may decrease the laxity of the submental tissues leading to a 'lift' of this area. In lateral photographs, the submental area appears smaller. Oni et al. [10] measured an average reduction of 45 mm 2 at 3 months. A lift could be seen in 56 of the 78 treated Caucasian patients. A double pass technique was used to treat the skin at two different depths (3.0 mm and 4.0 mm). Lu et al. [14] used the same method and found a reduction of 26 mm 2 after 3 months and 14 mm 2 after 6 months in Asian patients. Sasaki et al. [18] measured the displacement of the marionette lines 3 months after an MFU treatment. The lift varied between the treatment regimens from 2.4 to 3.8% along a line from the inferior tragal notch to the midpoint of the marionette line.

Subjective Measurements (Tables 4 and 5)
Different scoring systems were used to assess the subjective treatment results. The Investigator Global Aesthetic Improvement Scale (IGAIS, Table 4) rates the improvement or worsening in skin laxity on a 4-point scale [9,13,18,23] or a 5-point scale [12,14,19,24]. Both the 4-and 5-point IGAISs have the same four categories ranging from 'no change' to 'significant improvement', but the 5-point scale includes the item 'worsened'.
In the pooled IGAIS (n = 337), a total of 92% of the patients demonstrated an improvement at day 90; the 'mild improvement' category was greater (47%) than the 'moderate improvement' category (36%) ( Table 4). A shift was observed after the 180-day evaluation: Although the majority of patients still had 'mild' and 'moderate' improvement scores, the moderate category had increased (52%). 'No change' or 'worsening' was only present in a minority of patients (<7%). The patients could also self-evaluate the effect of the MFU treatment using the 4-point Subject Global Aesthetic Improvement Scale (SGAIS) ( Table 5). The pooled SGAIS (n = 81) was predominantly 'mild improvement', which increased from 42 to 53% in the period 90 to 360 days post-treatment. The 'moderate improvement' category also increased from 13 to 32% between the different time intervals. Worsening did not occur, whereas the 'no change' category decreased with time from 25% at 90 days to 5% at 360 days. Of note is that Fabi et al. [24] reported a significantly higher SGAIS score among patients with a BMI of ≤25 kg/m 2 . Oni et al. [10] reported less improvement in subjects with a BMI exceeding 30 kg/m 2 . Out of the 11 patients whose BMI exceeded 30 kg/m 2 , only three showed improvement.
Park et al. [11] used a self-modified photographic scale evaluating seven facial areas in 20 patients. Each area was scored from 0 (no wrinkles) to 4 (severe). A HIFU treatment decreased the wrinkle score by 0.9 points at 3 and 6 months, with the most effect on the jawline and cheek area.
Saket et al. [16] evaluated the MFU effect with an 'observation' scoring system ranging from 10 (no efficacy on wrinkles and lifting effect) to 100 (maximum efficacy). After 3 months post-treatment, most of the 22 patients treated scored between 58 and 66 (physician rated) or 50-60 (patient-rated).
Friedman et al. [20] treated lower face and neck laxity with ultrasound in 43 patients. Based on a global grading scale from 0 (exacerbation) to 5 (75-100% improvement), at three months, an improvement was seen in nine patients who had only slight sagging. The others showed no effect.
Araco et al. [21] used his self-developed Surgeon Assessment Scoring System scores and patient satisfaction questionnaire scores in 50 patients who had one MFU treatment.
After 6 months, the surgeon's score was 80, indicating a moderate effect on skin texture and a minimal-to-moderate face-lifting effect as visible in photographs. Patients indicated a 'moderate' to 'good' difference after treatment.

Devices, Treatment Regimens and Adverse Effects (Table 2)
Eleven studies used the Ulthera system (Ulthera, Mesa, Arizona, USA) [8,[12][13][14]17,18,21,24] and two used the Ultraformer system (Classys Inc., Seoul, Republic of Korea) [9,22]. The UTIMS A1 (Korust Co., Ltd., Seoul, South Korea) [16], the Doublo IFUS (Hironic, Yongin-si, Republic of Korea) [20] and the Microson (Microson, Cosmoplus Co., Sungnam, Republic of Korea) [19] devices were each evaluated once. An MFU treatment regimen consists of the systematic treatment of an area of skin area along different parallel treatment lines so that the multiple 'shots' are evenly distributed along these treatment lines in a specific area. The treatment can be performed at different tissue depths, typically 1.5 mm (dermal), 3 mm (subdermal) and 4.5 mm (the SMAS layer). The frequency of the ultrasound transducers ranges from 2 to 10 MHz, but the 4 and 7 MHz probes are used the most. The higher the MHz, the shorter the wavelength and the less deep the skin can be penetrated. The energy that accumulates in one spot (TIZ) ranges from 0.25 Joules (J) to 1.2 Joules, with the majority of settings ranging between 0.3-0.9 J. Multiple passes can be performed at different depths and directions. The total delivered energies can go up to 7200 J for full face treatment and 4600 J for the neck [18]. A total of 565 patients from 13 studies reported pain in all areas and at all device settings. Pain is rated on a 0-10 visual analog scale (VAS) or a numeric rating scale (NRS). The average reported pain score was in the 3.8 [2.5-6.1] range. The periorbital region and submandibular region treatments were more painful with higher VAS scores [10,18]. It should be noted that the pain was scored after applying topical anaesthetic ointment and/or systemic analgesics. Other side effects than pain were evaluated in a total of 573 patients. Transient erythema with or without oedema occurred in almost all the patients [13,14,18,22,23]. More uncommon were ecchymosis/bruising in four cases [9,11], transient dysesthesia in four cases [18,19], a wheal on the cheek in one case [10], skin burns in two cases [14], white linear striae of the neck in two cases [17] and one case of dermal white papules on the neck [24]. These uncommon adverse effects were noted in 2% of the total number of treated patients (14/573).

Discussion
This systematic review was undertaken to assess the effect of a single treatment of MFU on skin tightening. The overall results based on the IGAIS/SGAIS scores show that HIFU improved tightness of the skin to various degrees in most (>90%) of the included patients.
Clinically, the effect of a single MFU treatment results in long-term skin improvement. This can also be observed in the pooled IGIAS data (Table 5) with the percentage of patients in the 'moderate improvement' group increasing from 36% to 52% and the 'mild improvement' group decreasing from 47% to 34% during the 90 to 180 follow-up days. Support for this improvement is also seen in the pooled SGAIS data where the 'no change' category decreases from 25% at the 3-month, to 17% at the 6-month and 5% at the oneyear follow-ups. MFU treatment entails applying numerous subdermal small heating points with temperatures of about 60-70 • C [24], resulting in localised denaturation of the collagen [25]. The denatured collagen proteins are then gradually replaced by newly formed collagen fibres; this neocollagenesis leads to thicker and tighter skin [4,6,26].
The prolonged effect of an MFU treatment may also be due to the remodelling process which begins on day 28, as demonstrated by Hantash et al. [6] and Keagle et al. [27]. The expression of HSP47, a heat shock protein involved in wound healing through fibroblast proliferation and collagen production via a STAT3 signalling pathway blockade, becomes elevated 3 months after the heat treatment [5,27]. This expression of the HSP47 heat shock protein implies neocollagenesis and may reflect the reason for the improvement in skin laxity/tightening persisting up to at least three months post-treatment. No particular area of the face seems to be most susceptible to MFU, though Sasaki et al. [18] noticed a higher response in the brow region compared to the nasolabial region when measured objectively. One study noted the highest response in the cheek area [16], whereas another study did not see great differences between the regions when assessed separately [11]. It seems quite difficult to assess separate areas of the face since the facial areas are connected to each other and each area has a different muscle tone and the subcutaneous tissue and skin vary in thickness. These may all influence the outcomes.
Multiple variables can be altered during MFU therapy. Different transducers can be used with different frequency and energy settings. The wavelength (MHz) determines the amount of tissue penetration. A higher frequency ultrasound penetrates tissue less deeply. The energy (Joules) applied determines the amount of tissue heating and is responsible for the effect. The number of treatment lines and TIZs can be altered, as can the direction of the treatment lines (linear horizontal, vertical, criss-cross). It seems that more energy and more treatment lines at different depths [18] increase the effectiveness. An in-depth discussion of the different settings and their possible effects are beyond the scope of this review.
Since this review only evaluated the results of a single MFU treatment, it would be interesting to know if multiple MFU treatments with different time intervals may have an additional effect on skin tightening. In theory, additional TIZ could be chosen during a second procedure, in the untreated dermis or in the SMAS for additional skin tightening. However, the problem of subjective bias is evident when assessing aesthetic results, whereas objective measurements are limited, and, when a certain 'lift' is measured, it does not mean that this is visible to the human eye or that it fits a more aesthetically pleasing outcome. To overcome a subjective bias, various studies chose to assess the IGIAS with a blinded approach whereby the reviewers had to first identify the pre-and post-treatment photographs and, when the correct post-treatment photo was identified, an 'improvement' grade could be given [10,13,15,17,19,23]. Alam et al. [17] asked two or three reviewers to evaluate the photographs and then only rated a result as 'improved' if there was an agreement between the two reviewers. It would be interesting to see if an alternative to this subjective scoring system could be developed, perhaps with an aesthetic score using 3D image evaluations or the use of artificial intelligence.
Although MFU seems to be highly efficacious for skin tightening, there are limitations to these results.
First, the presence of excessive skin laxity was an explicitly mentioned exclusion criterion [8,10,12,20]. Three studies measured skin laxity severity and demonstrated that improvement declined with increased baseline skin [10,18,20]. This suggests that MFU treatment is not so effective in more severe skin laxity cases, and a surgical approach would be the best option for them. Secondly, the included patients were mostly adult females (>90%), and it would be interesting to know if the results can be reproduced in men since male skin is different from female [28,29]. Thirdly, the presence of excessive subcutaneous fat [10,12] or a high BMI > 30 [12] was often taken as exclusion criteria. Two studies [10,24] concluded that lower BMI values positively influence the outcomes. Fabi et al. [24] reported a significantly higher SGAIS for patients with a BMI of ≤25 kg/m 2 . Oni et al. [10] reported less improvement in subjects with a BMI exceeding 30 kg/m 2 . In a later study, Werschler et al. [12] excluded all patients with a BMI exceeding 30 kg/m 2 , enforcing the implicated negative influence of a high BMI on treatment outcomes. A reason why a higher BMI leads to less therapeutic effect may be that the skin tension of such cases may be higher due to more facial volume. This higher tension would then counteract the skin shrinkage following MFU.
In conclusion, MFU treatment seems to be effective for tightening the skin of patients with mild-to-moderate skin laxity. It may be less for those with a BMI > 30. This could be corroborated by future studies that should also focus on male patients, on optimising treatment regimens and on ways to score treatment outcomes more objectively.

Conflicts of Interest:
The authors declare no conflict of interest.