Steven Harris unveils a new classification to better describe and visually quantify a common issue when injecting filler into the lips

Steven Harris, MB BCh, MBCAM, MSc

One of the most popular nonsurgical procedures in the world is lip augmentation or enhancement with the use of hyaluronic acid (HA) fillers. Certainly on social media, they receive much attention and yet natural-looking results are very rare; the results we tend to see involve distortions of the natural lip shape with a degree of filler spread commonly referred to as filler migration.

In particular, the very popular ‘tenting’ or ‘fencing’ technique (including a slight variant, referred to as the ‘Russian lip’ technique) not only involves a relatively high risk of a vascular occlusion1 but creates an unnatural ledge. It does this by a series of perpendicular linear thread injections through the white roll of the vermillion border (VB), which flatten and distort the shape of the lips. The filler then backtracks along the multiple ‘channels of spread’ created by the needle to end up in the ‘white’ roll as an unnatural bulge, confusingly referred to as ‘crisp’ border.

Figure 1 Filler spread to form a plateau (A). The filler was dissolved and the lips treated using the Nonsurgical Lip Lift (NLL) which avoids direct injections of the vermillion border (VB).

A further exaggeration of injecting the vermilion border directly leads to the formation of a ‘shelf’, or in extreme cases a ‘plateau’, referred to as ‘duck’ lips (Figure 1). The ledge, shelf or plateau may be immediately apparent at the end of the lip procedure, erroneously referred to as ‘swelling’, or may progress over time (from a ledge to a shelf to a plateau). The process of filler spread not only creates unsightly, often offensive looking lips, but has an ageing effect on the face as a whole (Figure 2).

The Harris Classification

A new classification system, the Harris Classification of Filler Spread, was developed on the basis of an observational study. The purpose of developing a classification was to understand the different presentations of spread and gain an understanding of the underlying mechanism.
An observational study took place in North London between September 2019 and February 2020. It involved 48 Caucasian women aged 22–63 years, most of whom presented to the ‘in-house’ Dissolving Clinic for dissolving of filler spread, 2 weeks to 12 years after having had their treatments elsewhere (in the UK and abroad).

Figure 2 The ageing effect of filler spread. (A) Filler spread pulls the face forward simulating the effects of ageing which involve shrinkage towards the midline. (B) The Nonsurgical Lip Lift (NLL) avoids direct injections of the vermillion border (VB) to produce an anti-ageing effect.

The degree of filler spread was measured based on palpation (firmness) of the filler and visual inspection (unnatural protrusion and pale, grey or white discolouration) with further photographic evidence examined off-site. The unnatural protrusions were determined using Ricketts E-line and new V-lines developed by the author (Figure 3).

Four classes of filler spread were identified (Figure 4):

  • Class 1: No spread. The filler is placed in the superficial vermillion side of the border as in the Nonsurgical Lip Lift (NLL)
  • Class 2: Ledge. The filler forms an abnormal bulge of the ‘white’ roll. This is typically seen in the tenting/fencing technique or the Russian lip technique
  • Class 3: Shelf. The filler spreads beyond the ledge and may be readily apparent or progress over time
  • Class 4: Plateau. Here the filler spreads beyond the shelf.
    The filler spread was palpable in the majority of subjects as an area of firmness separate to the surrounding soft tissue. In all, there was a pale, grey or whitish discolouration with a clear demarcation from the natural, healthy-looking skin.

It is important to note that a minority of people do present with a natural ledge (which slopes rather than bulges), or shelf often seen with pouting, but these are not usually accompanied by the other changes seen with techniques involving spread; these changes include distortions of shape with abnormal projections, indentations, and unnatural light reflexes.

Figure 3 The V-lines. Line 1 is drawn from the columella-lip junction to the labiomental fold and down to the bony margin of the mandible; line 2 is drawn from here to the nasal tip. V1 is the measure of the lower lip projection ideally between the two lines (should not cross line 2) and should match the projection of the upper lip (V2) and chin (V3).

The underlying mechanism of spread

The results of the study support the view that filler spreads superficially most likely in the subcutaneous (superficial fat) layer along a path of least resistance pushed out further by the activity of the orbicularis oris muscle (OOM). The OOM consists of two distinct parts meeting at the vermillion border; the pars marginalis — a closed section contained within the vermillion, and the pars peripheralis — an open section in the cutaneous lip (Figure 5).

The pars marginalis forms a continuous band from modiolus to modiolus; it consists of mostly single band, narrow-diameter muscle fibres that meet at their medial end to interlace and attach to the dermis of the vermillion zone. The pars peripheralis is the thinner and more peripheral portion of the muscle. Its fibres decussate at the midline to insert into the contralateral philtral ridge, and they are reinforced by attachments from many of the regional muscles. Its most peripheral fibres are connected with the maxillary bone and nasal septum above and the mandible below. The posteriorly located pars peripheralis is described as consisting of horizontal, oblique and incisal (longitudinal) fibres, whereas the anteriorly located pars marginalis is described as having only horizontal fibres.2
Contraction of the pars peripheralis fibres produces labial elevation involved in both facial expression and speech. It is also responsible for the accentuation of perioral lines, referred to as ‘smoker’s lines.’ The pars marginalis fibres act primarily on the portion of the lip covered by the vermillion. These press the lip to the maxillary teeth or invert it closer to the oral cavity, wrapping the lip around the incisal and occlusal borders of the teeth. They are also involved in human speech for the production of labial sounds.3

Figure 4 The Harris Classification of Filler Spread. Class 1 represents no spread (normal) and classes 2-4 involve increasing degrees of spread; a ledge (class 2), a shelf (class 3) and a plateau (class 4).

The bulk of the lips are formed by the OOM which is in very close proximity to the surface and curls on itself (in a J-shape) so that the anterior projection of the pars marginalis gives rise to the ‘white skin roll’ of Gillies-Millard. The ‘white’ appearance is often not seen in darker skin colour and the section is more accurately referred to as the ‘skin roll’ of the lips, or the lip roll.4

Thus, filler placed in the superficial subcutaneous vermillion enters a closed system with contraction of the pars marginalis, keeping it closer to the oral cavity. When the filler is placed inside the vermillion border or lip roll, it enters an open system where it may easily spread outwards along a path of least resistance further pushed by the activity of the pars peripheralis and its supportive muscles. In this way, a ledge (class 2) spreads to form a shelf (class 3) and then a plateau (class 4) over time.

Figure 5 The orbicularis oris muscle (OOM) consists of the pars marginalis (A) and the pars peripheralis (B), which meet at the vermillion border. Spread takes place in the subcutaneous (superficial fat) layer (C). Image reproduced with permission from 3D4 Medical:

With age, the OOM loses its J-shape to form an I-shape; this plays a significant role in the lengthening of the upper lip and may explain the greater spread found along the thinning superficial layers in older patients, even when the filler is placed within the closed system of the vermillion.
Another possible mechanism of spread may involve injection into the muscle (pars marginalis) itself, which again would backtrack to end up superficially and spread from there with ongoing contraction of the muscles as described above. It is very unlikely that the filler would spread along the muscle itself without a path created by the needle (or a cannula) and the filler is readily visualized as a pale discolouration suggesting a more superficial (subcutaneous) spread in the perioral region.
Indeed, many other factors may play a role, such as interference with lymphatic drainage, variations in lip shape, and different configurations of the OOM. Filler rheology may have importance too, especially when placed at the vermillion border. Here hydrophilic fillers may spread more easily as may those with lower viscosity and cohesivity.

The tenting technique

Still, there can be little doubt that nonsurgical lip techniques involving direct injections of the vermillion border are the greatest cause of filler spread. The extremely popular ‘tenting’ (or ‘Russian lip’) technique is probably responsible for most cases of filler spread. It opens up the closed-lip system of the pars marginalis to the open system of pars peripheralis. The multiple perpendicular insertions of the needle at the lip roll, down to the centre do not only pose a greater risk of vascular occlusions but create ‘channels of spread’ where the filler backtracks to form the procedure’s signature ledge (class 2). Over time the filler spreads along the superficial fat layer pushed out by contraction of the pars peripheralis supported by the elevator muscles and the depressors in the case of the lower lip (Figure 6).

Figure 6 The ‘tenting’ technique involves multiple injections from the lip roll (white line) to the centre, forming channels of spread; the filler backtracks (black arrows) to form a bulging ledge and from here it is pushed out further (red arrows) by the contraction of the pars peripheralis.

While there are no studies to document the longevity of lip fillers in their spread position, it is the author’s observation that filler placed on the vermillion side of the border will tend to last up to a year; the ledge of class 2 lasts two years; class 3 persists for 3–5 years and class 4 for 5 years, or indefinitely. With increasing amounts of filler and degrees of spread, the filler appears to last longer.

Throughout this article, the word ‘spread’ has been used rather than the commonly used term ‘migration’. The latter is erroneously associated with certain brands of filler when the majority of cases can easily be attributed to poor injection techniques. Indeed, filler rheology may play a role once in the VB, but it is the technique used by the injector rather than the brand of filler which is responsible for it ending up there in the first place. As the filler does not have a mind of its own and cannot make the decision to ‘migrate’, the term ‘spread’ rightly places the responsibility on the injector who in most cases has the ability to make decisions.

The Nonsurgical Lip Lift

The Nonsurgical Lip Lift (NLL) was developed as an alternative to procedures involving direct injections of the VB; it consists of linear thread, curved thread, and bolus injections to work with the natural anatomy of the lips — their lines, curves, and tubercles in order to restore their ideal natural shape and rotation. All the injections are performed superficially on the vermillion side of the VB to gently press the soft side of the border against the hard one. As the lips rotate upwards, the philtrum is shortened and the philtral columns are indirectly defined too (Figures 7–8).
The NLL technique in classic form involves ten steps with no more than 0.05 ml of filler per step (Figure 8). The steps vary in terms of number and order depending on the presenting anatomy. As the VB is avoided, there is no migration of the filler so that rarely more than 0.5 ml is required for the entire procedure. Any reputable ‘soft’ filler (with low to medium viscosity and cohesivity) may be used and injections should ideally involve a needle as opposed to a cannula; the former allows greater control and precision.
It is the author’s opinion that in this particularly delicate area of the face aspiration should not be carried out. It is impossible to completely stabilise the syringe here so that during aspiration the needle may be inadvertently repositioned inside an artery to increase the risk of a vascular occlusion. In order to maximize safety, the steps involve only small amounts of product placed very gently, slowly, and superficially (no deeper than 2–3 mm below the surface). Common side-effects associated with lip treatments in general such as swelling and bruising (and the risk of more serious ones involving vascular compromise) are thus kept to an absolute minimum.
An integral part of the NLL involves the process of ‘tubercle shifts’. This allows the ‘shifting’ or reshaping of tubercles (with their crests and maximal light reflexes) in desired directions of movement. For example, injecting the lateral aspect of a tubercle will ‘shift’ it laterally (it’s crest and light reflex) to widen the relevant section (quadrant) of the lip. Adding filler to its medial aspect will ‘shift’ it medially to narrow the relevant lip section. Similarly, forward or backward shifting of the tubercle (along with the relevant lip section) may be achieved by injecting its anterior or posterior aspect respectively.

Figure 7 (A) Before and (B) after The Nonsurgical Lip Lift (NLL) using 0.3ml of filler.

Thus, the different tubercles may be recruited and shifted for the ideal lip shape.5

Figure 8 (A) The ten steps of the Nonsurgical Lip Lift involving bolus injections into the lateral tubercles (steps 1,5,6,10) linear threads (steps 2,4,7,9) with placement of filler in the central tubercle (steps 4 and 9) and curved threads (steps 3 and 8). (B) The H-planes (white lines) identify the ideal location of the tubercles while the H-curves (yellow curves) determine the optimal curvature of the lower lip in relation to the chin.

In many instances, the lip tubercles may not be readily identifiable and the practitioner may have difficulty determining their ideal position. While there are many measures to decide the ideal proportions of the lips such as the golden ratio and their position (developed by Steiner, Ricketts, Burstone, Sushner and Holdway)6, none specifically exist for the lip tubercles. In order to determine their ideal position, the author has developed H-planes (Harris planes, or H-lines) to locate the apex of each one on a frontal view (corresponding to their maximal light reflexes). Lip tubercles are naturally occurring soft dermal projections largely responsible for the shape of the lips and their natural light reflexes; there are in total five — three on the upper lip (upper right, upper left, and central) and two on the lower lip (lower right and lower left). Lines drawn from the top of the philtral columns at the base of the nose through the peaks of the Cupid’s bow will traverse the crests of the lower tubercles (at their points of maximal light reflexes). A line drawn down the middle (from the base of the nose) will cross the central tubercle at its crest. If the lateral H-lines are extended to the jawline, then the ideal curvature of the lower lip will match that of the chin; the author has named these H-curves or Harris curves (Figure 8). There appears to be greater variability in position of the upper lateral tubercles; however, shifting them towards the H-lines appears to lead to more aesthetic outcomes.


In conclusion, the all too common unaesthetic results from lip procedures tend to involve direct injections of the VB, in particular relating to the ‘tenting’ technique. Filler spread is immediately obvious as a palpable raised pale discolouration at the border or perioral region but may progress over time with the addition of more filler and/or the activity of the OOM. The latter is made of two parts; the pars marginalis and the pars peripheralis, which meet at the VB. Direct injections of the border open the closed system of the pars marginalis to the open system of the pars peripheralis. The degree of spread may be classified by the Harris Classification of Filler Spread into no spread (class 1), a ledge (class 2), shelf (class 3) and plateau (class 4). Newly introduced V-lines help measure the ideal projections of the lips and chin while H-lines and H-curves help determine the ideal positions of the tubercles and the curvature of the lower lip in relation to that of the chin. It is hoped that procedures such as the NLL will encourage further movement towards normal and natural looking results.

Declaration of interest None
Figures 1–4, 7, 8 © Dr Harris; 5, 6 © 3D4 Medical


  1. Cotofana, S., Pretterkliebar, B., Runhild, L., Konstantin, F. Distribution Pattern of the Superior and the Inferior Labial Arteries – Impact for Safe Upper and Lower Lip Augmentation Procedures. Plastic & Reconstructive Surgery, January 139(5):1. 2017
  2. Jain, R.P., Anatomy, Head and Neck, Orbicularis Oris Muscle. StatPearls (internet), January. 2020
  3. Rogers, C.R., et al, Comparative Microanatomy of the Orbicularis Oris Muscle between Chimpanzees and Humans: Evolutionary Divergence of Lip Function. Journal of Anatomy, January 214(1): 36-44. 2009
  4. Chandran, G., Lalonde, D.H., Obtaining a Good Lip Roll in Congenital, Secondary and Traumatic Cleft Lip Repairs. Canadian Journal of Plastic Surgery, Winter 21(4): 248. 2013
  5. Harris, S. The Nonsurgical Lip Lift (NLL) with tubercle shifts, H-lines and H-curves. Prime Journal, June, 12-15. 2020
  6. Joshi, M., Wu, L.P., Mahrajan, S., Regmi, MR. Saggital lip positions in different skeletal malocclusions: a cephalometric analysis. Progress in Orthodontics, May 16(1). 2015