Superficial versus deep injections of the upper midface—A prospective interventional split‐face study

Various injection algorithms have been proposed in the past which are in line with the three aesthetic principles: upper face first, lateral face first, and deep regions first. However, increasing evidence is provided that the upper midface can be targeted with superficial soft tissue filler injections alone too.


| INTRODUC TI ON
In a recent publication, the term facial biomechanics was introduced by Cotofana and colleagues. 1 The authors used this term to describe movements of facial soft tissues occurring during various facial expressions and during facial aging.3][4] This is in alignment with currently accepted treatment algorithms that predominantly target deep = supraperiosteal fascial layers to reshape and revolumize various facial regions including the midface.[7] Despite its global acceptance, the latter principle however (deep layers first) is being challenged by anatomic findings which investigated the superficial facial fat compartments individually and have reported that some superficial fat compartments move more (superficial nasolabial and jowl) whereas others are stable (superficial medial cheek and superficial lateral cheek) when experimentally tested and related to the underlying bone.Especially the superficial medial cheek fat compartment (which is also termed malar fat pad) seems to not undergo positional changes which might make this fat compartment also a target for superficial soft tissue filler injections in addition to the already established deep injections. 3om a safety standpoint, both planes can be considered "safer" injection planes because no major vessel can be found within the superficial fatty layer and in the deep plane, the small branches present arise predominantly from the external carotid artery without immediate connection to the ophthalmic artery circulation. 8om an aesthetic outcome standpoint, it might be interesting to evaluate which of the two targeted planes (superficial or deep) provide a better surface projection with consecutive improvement of upper cheek fullness and nasolabial fold severity if the same product and product volume is utilized.
Therefore, the objective of this study to investigate in a split-face study design the aesthetic outcome and the 3-dimensional local and regional effects of minimally invasive soft tissue filler injections of the upper midface if the superficial versus deep fascial planes are targeted using the same hyaluronic acid-based product and administering the same volume.

| Study sample
The investigated study sample consisted of 20 female volunteers of 100% Caucasian (Spanish) ethnic background.The mean age was 43.95 (11.9)This study was performed in adherence to the Declaration of Helsinki (1996), and in accordance with regional laws and good clinical practice for studies in human subjects. 11The study was conducted between September 2021 and June 2022, and all treatments were performed at REDACTED by the same injector to assure consistency during the treatment process.The study received ethics approval under the number of REDACTED.

| Study design
This study was designed as a prospective interventional split-face study with a total follow-up of 7 weeks and a touch-up interim possibility at 3 weeks.Study participants were treated on one side of the face (random assignment) with a deep injection, whereas the contralateral side was treated with a superficial injection utilizing for both approaches up to 1.0 cc of Belotero (R) Volume (Merz Pharma GmbH & Co. KGaA, Frankfurt am Main, Germany) as needed.
(Figure 1) After 3 weeks, the aesthetic outcome was evaluated and a touch-up with the same injection technique was applied if the aesthetic correction was incomplete or asymmetric based on the evaluation of the treating physician.At 7 weeks, the final aesthetic outcome was evaluated for each side separately.

| Deep injection technique
The deep injection technique relied on the administration of an average volume of 0.78 cc (0.2) of hyaluronic acid-based (HA) soft tissue filler via a 22G blunt-tip cannula (TSK Laboratory Europe, Oisterwijk, the Netherlands) via a sequential retrograde bolus injection targeting the supraperiosteal plane of the zygomatic arch and of the lateral midface (Figure 1).

| Superficial injection technique
The superficial injection technique relied on the administration of an average volume of 0.78 cc (0.2) of hyaluronic acid-based (HA) soft tissue filler via a 25G blunt-tip cannula (TSK Laboratory Europe, Oisterwijk, the Netherlands) via a sequential retrograde fanning injection targeting the subdermal plane of the zygomatic arch and of the lateral midface (Figure 1).
Following product administration, the correct plane (deep vs. superficial) of product placement was verified via ultrasound imaging.

| Clinical outcome assessment
Clinical outcome assessment was performed via pictures by four independent observers and for each facial side separately before the treatment, immediately after the treatment, before the touch-up injection at 3 weeks, and at the end of the follow-up period at 7 weeks.
The following outcome variable were evaluated: • "Crow's Feet at Rest Scale" (4-Point Likert scale ranging from 1 to 4; 1 = "no wrinkles", and 4 = "very severe wrinkles") 12 • "Upper Cheek Fullness Scale" (4-Point Likert scale ranging from 1 to 4; 1 = "full upper cheek", and 4 = "very severely sunken upper cheek") 12 • "Nasolabial Fold Severity Scale" (4-Point Likert scale ranging from 1 to 4; 1 = "no folds", and 4 = "very severe folds") 12 The aesthetic improvement was evaluated via the GAIS ("Global Aesthetic Improvement Scale"); a 5-Point Likert scale ranging from 1 to 5 with 1 = "Very much improved" and 5 = "Very much worsened".Clinical evaluation of the GAIS was performed in person for each facial side separately by the treating physician and the patient and was performed on patient images by the four independent observers.

| Objective outcome assessment
4][15] In brief, standardized fullface 3D images were obtained utilizing a Vectra H2 camera system (Canfield, Parsippany, NJ, USA).The images were automatically aligned to the baseline image and differences in volume (volume change in cc) and in skin position (skin vector displacement in mm) were calculated following the internal computational algorithm of the Mirror software tool kit for the middle, lower, medial, and lateral facial regions (Figures 2,3,4).

| Statistical analysis
Clinical scores, as assessed by the four observers, was averaged and the mean value was used for data presentation and statistical calculation.Comparisons between facial sides and between assessment points was performed by nonparametric paired testing using Wilcoxon signed-rank test due to the small sample size.All calculations were run using SPSS Statistics 25 (IBM, Armonk, NY, USA), and differences were considered statistically significant at a probability level of ≤0.05 to guide conclusions.

| Baseline comparisons
At baseline (before the treatment) no statistically significant difference was detected between facial sides for the crow's feet score (   Follow-up visits (3 and 7 weeks) revealed a statistically significant improvement in the crow's feet at rest score when compared to baseline with all p ≤ 0.002 for both deep and superficial injections (Figure 5).

| Upper cheek fullness scale
Immediately after the treatment, the score changed from its baseline value (see above) to (deep vs. superficial) 2.39 (0.4) versus 2.41 (0.4) with p = 0.317.After 3 weeks follow-up the scores were (deep vs. superficial) 1.94 (0.3) versus 1.90 (0.4) and p = 0.429 indicating no statistical difference between facial sides.After 7 weeks follow-up the scores were (deep vs. superficial) 1.67 (0.3) versus 1.59 (0.3) with p = 0.109 indicating no statistical difference between facial sides.Follow-up visits (3 and 7 weeks) revealed a statistically significant improvement in the upper cheek fullness scale when compared to baseline with all p < 0.001 for both deep and superficial injections (Figure 6).

F I G U R E 5
Bar graphs showing the "crow's feet at rest scale" (0-4, best to worst) as evaluated by independent observers before treatment, after treatment, at 3 weeks follow-up, and at 7 weeks follow-up.Scores are compared between deep (orange) and superficial (blue) injections.

F I G U R E 6
Bar graphs showing the "upper cheek fullness scale" (0-4, best to worst) as evaluated by independent observers before treatment, after treatment, at 3 weeks follow-up, and at 7 weeks follow-up.Scores are compared between deep (orange) and superficial (blue) injections.

| Lower facial volume
Immediately after the treatment, the lower facial volume decreased following the deep versus superficial injections by −0.11 (0.9) cc versus −0.02 (1.0) cc with p = 0.478.After 7 weeks follow-up the lower facial volume decrease was for the deep versus superficial injections −0.10 (1.1) cc versus −0.05 (1.0) cc with p = 0.837 (Figure 9).

| Medial midface repositioning
Immediately after the treatment, the vertical skin vector displacement for the medial midface was for the deep versus superficial injection −0.05 (0.5) mm versus −0.01 (0.4) mm with p = 0.841.After 7 weeks the vertical skin vector displacement for the medial midface was for the deep versus superficial injection −0.04 (0.5) mm versus −0.22 (0.6) mm with p = 0.126 (Figure 10).

| Lateral midface repositioning
Immediately after the treatment, the vertical skin vector displacement for the lateral midface was for the deep versus superficial injection 0.08 (0.7) mm versus −0.07 (0.8) mm with p = 0.334.After 7 weeks the vertical skin vector displacement for the medial midface was for the deep versus superficial injection −0.03 (0.7) mm versus −0.52 (0.9) mm with p = 0.723.(Figure 11).

| Patient GAIS
The aesthetic improvement following the treatment as assessed at 7 weeks by the patients on a scale from 1 to 5, best to worst was for the deep injections 1.74 (0.9) whereas it was for the superficial injections 1.84 (1.0) with p = 0.671.

| Observer GAIS
The aesthetic improvement following the treatment as assessed at 7 weeks by the four observers on a scale from 1 to 5, best to worst was for the deep injections 2.04 (0.6) whereas it was for the superficial injections 2.04 (0.4) with p = 0.902.

| Treating physician GAIS
The aesthetic improvement following the treatment as assessed at 7 weeks by the treating physician on a scale from 1 to 5, best to worst was for the deep injections 2.16 (0.8) whereas it was for the superficial injections 2.21 (0.8) with p = 0. 317.

| Adverse events
No adverse events related to safety or product tolerability were observed during the entire study period.

| DISCUSS ION
The results of this prospective, interventional split-face study revealed that both injection techniques (superficial and deep product administration) have provided safe aesthetic treatments; no adverse events related to safety or product tolerability were observed during F I G U R E 7 Bar graphs showing the "nasolabial fold severity scale" (0-4, best to worst) as evaluated by independent observers before treatment, after treatment, at 3 weeks follow-up, and at 7 weeks follow-up.Scores are compared between deep (orange) and superficial (blue) injections.the entire study period.Despite being a split-face study, which administered the product in two different fascial planes (superficial vs. deep), no statistically significant side differences were observed at the end of the 7 weeks observational period when the aesthetic improvement was rated via the GAIS by the patient (p = 0.671), by the four observers (p = 0.902), or when rated by the treating physician (p = 0.317).This indicates that despite the product was administered in two separate planes the aesthetic outcome showed no statistically significant differences independent of who rated the outcome after 7 weeks.However, it has to be noted that the outcome was evaluated in repose and not under dynamic conditions like smiling. 16sessing the outcome also under dynamic condition could have F I G U R E 8 Bar graphs showing volumetric changes in the midface (in cc) as measured in the 3D analysis software between before and after treatment and between before treatment and at 7 weeks follow-up.Scores are compared between deep (orange) and superficial (blue) injections.

F I G U R E 9
Bar graphs showing volumetric changes in the lower face (in cc) as measured in the 3D analysis software between before and after treatment and between before treatment and at 7 weeks follow-up.Scores are compared between deep (orange) and superficial (blue) injections.
F I G U R E 1 0 Bar graphs showing skin repositioning in the medial face (in mm) as measured in the 3D analysis software between before and after treatment and between before treatment and 7 weeks follow-up.Scores are compared between deep (orange) and superficial (blue) injections.
potentially indicated a difference between superficial and deep product administration which was not conducted for this study; this must be regarded as a limitation of this investigation.
When analyzing the outcome after 7 weeks following the initial treatment, it was revealed that independent of the plane initially targeted (superficial vs. deep) all semiquantitative scores (Crow's feet at rest scale, Upper cheek fullness scale, Nasolabial fold severity scale) statistically significantly improved with all p < 0.001.This indicates that the cannula injection techniques performed are effective in improving the locale volume (= Upper cheek fullness scale: from 2.8/2.9 to 1.7/1.6)but have also regional effects.This can be seen in the correction of the Crow's feet at rest scale (from 2.5/2.6 to 1.8/1.9)and in the improvement of the Nasolabial fold severity scale (from 2.5/2.5 to 1.8/1.8).The regional effects can be explained by the layered arrangement of the facial anatomy. 17The fascial layers of the upper and lateral midface (where the product was initially administered) are connected with each other via the 3-dimensional superficial musculo-aponeurotic system (SMAS) which can transmit soft tissue changes to another region even if those regions are apart. 18,19Injections along the zygomatic arch are located lateral to the line of ligaments 5,6,20 and can therefore precondition and thus improve medially located facial regions like the nasolabial fold even if the nasolabial fold was not targeted a priori. 1 The improvement of the lateral orbital region can be explained by the fact that the orbicularis oculi muscle is integrated into the SMAS and therefore changes to the SMAS affect the tension between the muscle and skin resulting ultimately in reduced wrinkle severity in the circumference of the lateral orbit. 21 addition to the performed multiobserver semiquantitative outcome scoring, objective facial regional analyses were conducted via 3-dimensional photography. 6,13The results revealed that the increase in midfacial volume when compared to the facial status before the treatment was (deep vs. superficial) 2.19 cc versus 1.94 cc at the 7 weeks follow-up.This is remarkable because only 0.78 cc of product were administered but the volume increase was 181% for the deep and 149% for the superficial injection.This can be explained by the rheologic properties of the product.The product is able to bind water due to its specific water binding abilities and thus increase the local volume of the infiltrated midfacial soft tissues. 224][25] Elevated soft tissues form a tent-like formation around the product, and this allows for a greater surface projection that just the volume of the administered product itself. 26This product behavior is supported by the fact that the increase in surface projection was observed immediately after the treatment and continued until the 7 weeks follow-up period.It can be speculated that this effect might continue even further but future studies need to investigate this trend specifically with a longer follow-up period.
Despite not statistically significant (p = 0.616 and p = 0.332), a slight trend was visible in which the deep injection effected a greater surface projection than the superficial injection technique.This is plausible because the deep injection technique elevates more soft tissues than the superficial technique; this can be compared to the bottom diameter of a large (= deep injection technique) versus a small (= superficial injection technique) tent: a larger tent has a greater overall surface than a small tent.This effect is additionally supported by volume measurements conducted in the lower face which have revealed that midfacial volume increase promotes lower facial volume decrease.This is in line with a previous publication 27 and provides an explanatory model for the effects seen in the lower

| CON CLUS ION
The results of this split-face study reveal that both the superficial and the deep cannula injection technique for midface volumization statistically significantly improve the midfacial volume, reduce nasolabial fold and crow's feet severity.Both injection techniques were rated by the patient, independent observers, and the treating physician to be without side differences and to improve the overall aesthetic appearance of the face.No statistically significant difference was observed between the two injection techniques when compared via semiquantitative and objective outcome evaluation.Each author has participated sufficiently in the work to take public responsibility for appropriate portions of the content and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

FU N D I N G I N FO R M ATI O N
This study received financial support by Merz Pharma España S.L.

CO N FLI C T O F I NTER E S T S TATEM ENT
years [range: 28-65], and the mean body mass index (BMI) was 22.92 (2.7) kg/m 2 [range: 20.02-26.89].Study participants were recruited for this study via online advertisements and adhered to the following inclusion criteria: female gender, age between 25 and 65 years, Fitzpatrick skin type I-IV, body mass index between 20 and 27, bilateral, symmetrical, moderate-to-severe midfacial volume loss, 9 moderate-to-severe jawline contouring loss, 10 desire for volume loss correction, fully vaccinated for SARS-CoV2 at least 2 weeks prior to the start of the study, and willingness to comply with the treatment algorithm and the follow-up visits.Prior to the inclusion in this study all participants had to provide written and informed consent for the use of their demographic and imaging related data.Patients were not included in this study if they were pregnant or breastfeeding, had reported hypersensitivity to hyaluronic acid or lidocaine/anesthesia, had active skin disease (e.g., infection, eczema, dermatitis, psoriasis, acne, and rosacea) in the treatment area (= midface), had autoimmune disorder or cancer (unspecified), previous treatment in midface with nonpermanent soft tissue filler and/or neuromodulators in last 12 and 6 months, respectively, had previous facial surgery (unspecified), had scars or permanent-or semipermanent fillers in their midface, planned to undergo any other aesthetic treatment in the midface during the duration of the study, or had COVID-19 symptoms or a confirmed positive infection.

F I G U R E 1
Illustrations showing the deep and superficial injection technique with the respective entry points and injection points utilizing a cannula.3.1.2| Touch-up evaluations After the 3 weeks follow-up evaluation, 60% (12 out of 20 study participants) of those treated with the deep injection technique required a touch-up treatment with an average volume of 0.24 (0.1) cc, whereas 40% (8 out of 20 study participants) of those treated with the superficial injection technique required a touch-up treatment with an average volume of 0.24 (0.1) cc.Despite a difference in frequencies was observed (60% deep vs. 40% superficial) no statistically significant difference was detected when performing a chisquare test with p = 0.206.

F I G U R E 3
3D images of a 56-year-old female study participant showing the change of volume at 7 weeks follow-up when compared to baseline.She received a superficial injection on the left side and a deep injection on the right side for both the initial and the touch-up injection.F I G U R E 2 3D images of a 29-year-old female study participant showing the change of volume at 7 weeks follow-up when compared to baseline.She received a superficial injection on the left side and a deep injection on the right side for both the initial and the touch-up injection.3.2.3| Nasolabial fold severity scale Immediately after the treatment, the score changed from its baseline value (see above) to (deep vs. superficial) 2.28 (0.6) versus 2.40 (0.6) with p = 0.039.After 3 weeks follow-up the scores were (deep vs. superficial) 2.09 (0.5) versus 2.09 (0.6) and p = 0.957 indicating no statistical difference between facial sides.After 7 weeks followup the scores were (deep vs. superficial) 1.83 (0.4) versus 1.83 (0.4) with p = 1.00 indicating no statistical difference between facial sides.
Follow-up visits (3 and 7 weeks) revealed a statistically significant improvement in the upper cheek fullness scale when compared to baseline with all p < 0.001 for both deep and superficial injections (Figure7).

F I G U R E 4
3D images of a 56-year-old female study participant showing the skin vector displacement at 7 weeks follow-up when compared to baseline.She received a superficial injection on the left side and a deep injection on the right side for both the initial and the touch-up injection.
face.The volume decrease in the lower face was (deep vs. superficial) −0.10 cc versus −0.05 cc after 7 weeks follow-up with p = 0.837 between facial sides.This indicates that the decrease in lower facial volume was slightly larger for the deep injection which is supported by the fact that more facial soft tissues were repositioned with the deep versus the superficial injection technique.A similar effect was observed when analyzing objective skin vector displacement which is also in line with the results presented above.A deep injection can elevate the entire soft tissue envelope as a coherent unit whereas the superficial injection technique can cause a disconnection between the fascial layers separating superficial from deeper fascial planes.This is most likely caused by the administered product which creates a new layer between the F I G U R E 11 Bar graphs showing skin repositioning in the lateral face (in mm) as measured in the 3D analysis software between before and after treatment and between before treatment and 7 weeks follow-up.Scores are compared between deep (orange) and superficial (blue) injections.superficial and the deep structures.This disconnect reduces the connection of the superficial structures to the deep structures and affects thus the stability of the superficial layers.As a consequence, the superficial layers are more amenable to the effects of gravity and can displace caudally.This skin surface displacement into more caudal locations is confirmed by the greater magnitude of caudally pointing (= negative) skin displacement vectors which are greater for the superficial than for the deep injection technique (deep vs. superficial): medial midface −0.04 mm versus −0.22 mm with p = 0.126; lateral midface: −0.03 mm versus −0.52 mm with p = 0.723.Despite not reaching statistical significance, the results are coherent and point toward a trend.
AUTH O R CO NTR I B UTI O N S J.M.M., J.M.R., N.M.A., L.M.J., J.B.O., M.A.G., M.A., and S.C. have made substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data.J.M.M., J.M.R., N.M.A., L.M.J., J.B.O., M.A.G., M.A., and S.C. have been involved in drafting the manuscript or revising it critically for important intellectual content and given final approval of the version to be published.

M
.A. and S.C. declared no conflicts of interest with respect to the research, authorship, and publication of this article.All other authors provide consulting and speaking services for Merz Pharma España S.L. and/or its affiliates, and use Merz Aesthetics products in their offices.