Ultrasound Gives Us a New Understanding of Filler Behavior and Spread

Ultrasound is becoming increasingly adopted in the field of aesthetics. It provides real-time images without radiation and is extremely portable and cost-effective. For the injector, it provides information regarding a patient’s unique anatomy and prior filler treatments, and helps elucidate complications. It is an invaluable tool to assess filler and is being used by researchers to better understand its behavior. 

One of the most frequently discussed complications on Instagram is “filler migration.” In this regard, lip filler has been the most implicated source of consternation among clinicians. Dr. Steven Harris has stated, “Filler doesn’t just pack its suitcase and decide to move to another location” (S. Harris, MD, oral conversation, 2026). The fundamental concept that the injector must embrace is that filler follows a path of least resistance, always. 

UNDERSTANDING FILLER REDISTRIBUTION

It has been widely accepted that when filler is placed in the dermis of the nasolabial fold, most of the material is found in the immediate subcutaneous layer by histologic evaluation.1 The research attributes this to the thinness of the dermis being < 1.5 mm.¹ In a trial performed in my office, it was found that 3 different fillers (SKINVEVE by JUVEDERM, Allergan Aesthetics; Restylane Defyne, Galderma; RHA 3, Revance Aesthetics) behaved similarly and preferentially flowed to the subcutis rather than remaining in the dermis during real-time ultrasound evaluation (S. Weiner, unpublished data, 2025). Over subsequent months, almost no filler was identified in the dermis using ultrasound. 

Although dermal thickness is a contributing factor, the primary cause is more likely that the dermis is not receptive to expansion and that filler finds the subcutaneous layer less resistant. 

Typical 30° to 45° needle insertion angles were shown to place filler almost always into the subcutaneous tissue in a study by Micheels et al published in 2016.² In contrast, insertion angles of 7° to 12° were needed to consistently place filler within the dermis. 

However, there is little margin of error with dermal injections. Remaining too superficial increases the risk of visible blebs, whereas slightly deeper “dermal” injections almost invariably deposit filler within the subcutaneous tissue. 

DISTRIBUTION WITHIN THE NASOLABIAL FOLD

When nasolabial fold injections were evaluated over several months, another phenomenon was found (S. Weiner, unpublished data, 2025). The filler gradually redistributed laterally into the superficial nasolabial fat pad. The exact problem the filler was injected to correct was being exacerbated by volumizing the nasolabial fat pad. 

Why was this happening? The area medial to the nasolabial folds, the ergotrid, is composed of very compact superficial fatty layers, which are resistant to filler, whereas in the tissues lateral to the nasolabial fold, the superficial fat is more receptive to filler due to a less dense fibrofatty layer. This was characterized by changes in the SMAS on either side of the nasolabial fold.³ Another factor is that during expression, the lip elevators pull the nasolabial fold laterally, which would also exert a force on the filler to move in the same direction. 

EXPLAINING LIP MIGRATION 

Returning to the concept of lip migration, there is a straightforward explanation for these findings. Filler will spread or redistribute along tissue planes in a path of least resistance. If a lip is overfilled, the lip compartment eventually becomes unable to withstand further expansion. Consequently, filler preferentially redistributes superiorly beyond the vermilion border or inferiorly beyond the wet-dry border. 

There is no anatomic compartment separating the lip from the ergotrid. Therefore, even relatively large filler volumes placed along the vermilion border tend to redistribute into the ergotrid.⁴

When the intended plane is the subcutaneous tissue, many clinicians inadvertently deposit filler within the orbicularis oris muscle. Cannula-based and “tenting” techniques appear more likely to lead to intramuscular filler placement.

NEEDLE BACKFLOW AND PLANE CHANGES

The concept of backflow leads to filler redistribution during injections. Whenever an instrument enters through the skin, a potential space is created around it. This space is approximately 33% larger than the diameter of the needle used and approximately 90% larger than the diameter of the cannula. Consequently, this tract becomes a path of least resistance for filler distribution. 

As a result, 60% of needle injections result in filler changing tissue planes.⁵ For example, a deep periosteal bolus may ultimately be deposited within the superficial fat rather than the intended deep plane. In contrast, despite the larger potential backflow tract associated with cannulas, no change in tissue planes was observed because of greater length of the cannula.5 

RETINACULA AND TISSUE PLANE DYNAMICS

A newly published paper describes a superficial retinacula called “the retinacula superficialis,” which spans between the SMAS layer and the skin, and a deep retinacula, called the “retinacula profundus,” which spans between the bone and the SMAS layer.⁶ The paper further elucidates that when filler is injected into the deep or superficial fatty layer, the retinacula helps direct the spread of filler within that tissue plane. Orientation of these fibers determines the predominant direction of filler spread. 

Although the SMAS layer is not permeable to filler placed on either side of it, filler does flow freely within the deep fat compartments. Again, the tissue plane offering the least mechanical resistance largely determines the pattern of filler redistribution. 

FILLER SPREAD ALONG THE ZYGOMATIC ARCH

Additional research found a very predictable pattern of filler spread when placed on the zygomatic arch.⁷ Ultrasound evaluation showed filler placed on the periosteal layer would immediately redistribute to the intermediate fat pad (superficial fat pad). The more filler injected along the zygomatic arch, the more extensive filler spread was observed. Over time, when this injection was repeated, negative aesthetic implications came with a widening of the temporal region. 

CONTINUITY OF THE DEEP FAT COMPARTMENTS

The newest edition of the Cotofana Anatomy textbook explains the evolving understanding of the deep fat pads of the upper and midface.⁸ They are all continuous, so spread of filler is possible to the suborbicularis oculi fat (SOOF), medial and lateral deep cheek fat pads, and intermediate temporal fat pad. There are no compartments of the deep fat pads as previously believed. 

Injectors must recognize that filler can redistribute freely throughout the deep fat compartment. Accordingly, caution should be exercised when attempting to overcorrect volume deficits within these anatomic spaces. 

PERIOSTEAL INJECTIONS IN THE TEMPLE, JAWLINE, AND MIDFACE

One of the most popular enhancement techniques in the temples and jawline is placement of a bolus of hyaluronic acid (HA) filler onto bone/periosteum. Injectors must understand that regardless of the injection angle, filler frequently redistributes into the adjacent muscle. Also, the initial result gradually diminishes over time as the filler redistributes throughout the temporalis and masseter muscles during normal activities such as chewing and speaking (Figures 1 and 2). 

With the temple region, there is also a resistive force created by the dense deep temporal fascia, which counteracts the initial tissue expansion produced by filler placement. To correct the aging midface, a common treatment is to place a bolus of filler on the periosteum in the deep pyriform space. This space is bounded superficially and superiorly by the levator labii superioris alaeque nasi (LLSAN) which goes superiorly to its origin near the tear trough. Laterally, it is continuous with the deep medial fat pad and inferomedially by the depressor septi nasi.⁹ Injections in this space will redistribute superiorly and medially as these are the path of least resistance. Overfilling can affect the natural movement of the LLSAN during expression and lead to abnormal dimpling.

CHIN PROJECTION AND FILLER REDISTRIBUTION

Deep chin injections for projection are widely performed. However, when needles are used, the concept of backflow remains ever present. Although the injector may intend to deposit filler directly on the periosteum, backflow frequently results in redistribution into the mentalis muscle and subcutaneous tissue during injection (Figure 3). 

Research on the anatomy of the retinaculum in the chin shows filler will preferentially flow along these fibrous bands and become mainly situated in the subcutaneous layer.⁶ This observation is further supported by an unpublished ultrasound study in which the author participated using a highly viscous HA filler (S. Weiner, unpublished data, 2024). It was hypothesized that due to the “thickness” and high G’ of the filler, the bolus would stay on the periosteum, but that the ultrasound findings did not support that hypothesis. If the filler is injected directly into the median raphe, a fibrofatty structure, it appears the filler will spread less. Injecting with a cannula, which pierces into this structure, under ultrasound guidance may represent the preferred technique for minimizing filler spread. Ultimately, the chin will widen as the filler spreads superficially and laterally along the retinacula network. 

PREAURICULAR REGION AND TEMPORAL FAT PAD CONSIDERATIONS

When injecting in the preauricular area, clinicians must be extremely cautious. The intended target is the subcutaneous fat. However, it may be 1 mm to 4 mm in thickness.¹⁰ It is very possible to inject into the parotid gland, with no abnormal sensation from the patient. Although most inadvertent parotid injections do not result in clinically significant sequelae, complications including infection, sialocele, xerostomia (dry mouth), and pain may occur. Ultrasound-guided injections can reduce the likelihood of this occurring.¹⁰

There is a projection of the buccal fat pad (Bichat fat) that extends into the temple and is known as the temporal fat pad. During deep temporal injections, filler may inadvertently be deposited within this fat pad. Once injected, filler may redistribute inferiorly through the fat pad into the lower face. 

Another interesting finding from ultrasound observation is that opening the jaw, which is a technique used by many clinicians during deep temporal injections, leads to the fat pad 

moving superiorly. This movement may increase the likelihood of inadvertent filler deposition within the fat pad. Therefore, this technique of jaw opening during injection should be avoided.

NONSURGICAL RHINOPLASTY

Nonsurgical rhinoplasty using HA filler has become a popular aesthetic procedure. As in the temple and jawline, the projection obtained immediately after injection will gradually decrease over time, as filler spreads more laterally. Unfortunately, widening of the radix of the nose can occur over time which can create an undesirable aesthetic result.

PRACTICAL IMPLICATIONS FOR CLINICAL PRACTICE

The concept of filler redistribution or spread is based on anatomy, injection volume, and time. A clinician must not fight this phenomenon but instead embrace it and plan injections accordingly. Remember that filler, regardless of type or manufacturing process, has the potential to redistribute along tissue planes that offer the least mechanical resistance. The concept of blaming the filler for “migration” should be abandoned; injector technique, injection plane, treatment volume, and facial anatomy are the primary determinates of filler redistribution.