Review of Botulinum Toxin Uptake and Novel Theory Regarding Potential Spread Days After Injection

Abstract

 

Although most clinicians agree that diffusion of botulinum neurotoxin (BoNT) occurs, its extent, timeline, and clinical importance have been debated. A literature search on PubMed (National Institutes of Health, Bethesda, MD) was done up to January 15, 2023, which included the following search terms: botulinum toxin A uptake; botulinum toxin A diffusion; and botulinum spread. A total of 421 publication titles were found and analyzed. Based on titles, the author selected 54 publications as potentially applicable and reviewed each 1 in depth, along with its supporting references. There are several publications that support a novel theory that small amounts of BoNT can remain in the area days after injection and spread to adjacent muscles. Although current thinking is that BoNT is completely taken up within hours, making its spread days after injection of BoNT not a plausible theory, the following literature review and case report support a novel theory.

Level of Evidence: 5

Although the mechanism of action of botulinum neurotoxin (BoNT) is well known, including its binding, translocation, and lysis, what happens to any remaining BoNT not taken up into the neuromuscular junction (NMJ) is not fully understood. Current thinking comes from work done by Simpson in 1980. He analyzed toxin added to rat hemidiaphragm tissue baths and reported that 50% of binding is complete at 12 minutes, and 90% of binding is complete at 40 minutes, and 50% of translocation is complete in 4.9 minutes.¹ Colasante et al later reported that in their experiment in mice “the majority of neurotoxin molecules are inside small clear synaptic vesicles in muscle samples fixed 5 or 10 minutes after injection.”² Authors in subsequent papers quote these studies and generalize that binding of BoNT occurs in “approximately 30 minutes,” or “within a few minutes, but . . . at most a few hours,” and that “most BoNT is inside synaptic vesicles in muscles 5 or 10 minutes after binding,” without explaining the context of in vitro animal studies and thereby implying that these statements can be adapted to live humans when they are injected with BoNT for cosmetic purposes in much smaller doses.^3-5 That animals and humans are not directly comparable is demonstrated by Hughes and Whaler, who utilized the same rat hemidiaphragm preparation bathed with toxin and reported the disappearance of response to direct stimulation at 30 to 60 minutes.⁶ This quick timeline of muscle paralysis is clearly not true of muscles injected with BoNT for cosmetic use in humans, which suggests that the binding and translocation times in these experiments also cannot be extrapolated. Further, a paper in 2000 by Huang et al quotes a paper by Yamada et al stating that binding of BoNT takes an “average of 32 to 64 minutes”; however, this paper was about aminoglycoside antibiotics and their effects on the neuromuscular junction.^7,8 Despite this, additional papers go on to quote Huang et al, repeating that BoNT “uptake takes only 32 to 64 minutes,” further confounding the literature.^9,10

Although understanding of the human body often starts with in vitro research studies or in vivo studies in animals, as discussed above, these results do not always correlate to humans receiving cosmetic doses. In these studies, there is immediate placement of the BoNT, bathing the analyzed muscle instead of diffusion from one injection point, and these baths do not have exposure to additional cell types or a gradient potential from neighboring tissues that could supply the NMJ with additional BoNT once the nearby BoNT is internalized. Additionally, a lot of these studies discuss the “majority” of BoNT without giving exact amounts, which is deceptive, because a minority of BoNT may remain locally and have a residual effect. Finally, animals and humans have very different responses to BoNT. For instance, although the half-life for elimination is several hours in mice, in humans it may be several days.^11-13 For these reasons, the assumptions made in many of these papers are not applicable to cosmetic injections done in the office.

In fact, Simpson, who is one of the most notable and widely cited investigators in this area, mentions that “large portions of BoNT can remain in the extra-cellular compartment and there does not yet exist a full understanding of the mechanism for metabolism and elimination of toxin in either the extracellular or intracellular compartments,” and continues by saying that the extravascular extracellular fluid compartment is the “largest fluid compartment to which the toxin is distributed.”¹¹ Although the BoNT in the extracellular space adjacent to cholinergic nerve endings will be taken up quickly, there is no evidence regarding the BoNT in the extracellular space where there are no target cells. Simpson also explains, in his 2013 paper, that his earlier experiments and many others were done on neuromuscular preparations” and although they showed that the half-life for toxin extraction from the “peri-neuronal compartment is relative rapid,” the perineuronal microcompartment was “miniscule compared to the total fluid volume of the extravascular compartments.”¹¹ Therefore, he hypothesized that the efficient extraction of toxin from this small area had only a minimal effect on the half-life for BoNT elimination from the entire fluid compartment.

Early studies examining binding and internalization were often focused on the nerve terminal of the muscle and did not have the capability to explore any other locations.^1,2 However, in an animal study by Tang-Liu et al, distant sites were examined after injection of radiolabeled BoNT in rats, and and it was found that the percentage of BoNT increases in the skin from 0 to 6 hours (0% to 25.4%), at which point it then begins to decrease at a very slow rate, whereas in the muscle the percentage of BoNT decreases after injection (from 97.7% to 18.7% over 0 to 6 hours).¹⁴ One could speculate that the BoNT was taken up in the muscle because it was placed intramuscularly next to its target cells. However, the BoNT in the skin required time to diffuse to this site's target cells, which were cholinergic eccrine and sebaceous glands. Interestingly, free iodide also was injected, and at every time point very minimal amounts (less than 1%) remained at the site of injection, and although free- ¹²⁵I-iodide was washed away from the extracellular fluid with none detected after 24 hours, that was not the case with the radiolabeled BoNT. Most importantly, radioactivity was found in significant amounts outside the injection site days after injection. At 48 hours, when only 1.46 to 1.84% remained in the muscle, 4.6% to 6.2% of I-BoNT/A was still present in the skin, supporting our hypothesis that a significant amount of active BoNT outside target cells may be available to create an effect days after injection.

Additionally, there are several studies examining the diffusion potential of BoNT in both animals and humans. In animals, Borodic et al showed that if doses were high enough, diffusion of BoNT could be up to 45 mm from the site of injection, with only 15 to 30 mm when lower doses where used, and Arezzo and Yaraskavitch et al also showed a dose-dependent spread to nearby noninjected muscles.^15-17 In humans, Eleopra et al showed a significant change in untreated muscles, which the investigators concluded was related to local diffusion of the BoNT.¹⁸ These studies all contradict Hallett's assumption, which is currently shared by many physicians, that “toxin that is not taken up into the nerve endings will be washed out,” likely “in minutes,” and will go into the general circulation.⁴ Although there also are studies that do not report any affect on neighboring muscles, it is likely due to a difference in other variables such as the total dose, concentration, or volume of BoNT; concentration of the target area; distance to the target area; muscle activity; and injection technique, such as speed, size of needle, placement, and number of injections.^19-24 There are several human studies that have reported that BoNT spread is determined by the fluid volume, 1 showing that a fivefold increase in volume resulted in a 50% increase in the affected area. In addition, in a review paper, Ramirez-Castaneda writes that “the spread of BoNT is dependent on a number of factors, one of the most important is the volume.”²⁵ The studies mentioned above showing a significant amount of BoNT in animal skin 24 to 48 hours after injection, the long half-life of BoNT in vivo, and BoNT's ability to diffuse farther with larger volumes of fluid all support the plausibility of this novel hypothesis.

Finally, most recently, a report by Dal’Forno presented a series of “unexpected BoNT-A adverse events . . . in patients who had BoNT-A treatment concomitantly with inflammation-inducing procedures or skin conditions.”²⁶ This report included 4 patients with upper lip ptosis, 2 of whom had same day laser resurfacing, 1 who had a chemical peel, and 1 who had photodynamic therapy (PDT), all with same-day injection of BoNT, and with significant facial swelling. The following case report similarly presents a patient with upper lip ptosis, likely due to spread from the glabella to the levator labii superioris alaeque nasi (LLSAN). The patient had large amounts of edema from laser resurfacing in the tissue that had been treated 2 days previously with BoNT. In light of recommendations to pretreat areas undergoing resurfacing with BoNT first, further investigation into optimal timelines for this pretreatment are necessary.²⁵ Further studies are planned to help elucidate the optimal timelines.

PATIENT CASE

A 66-year-old patient had a BoNT treatment (onabotulinumtoxinA reconstituted with 2.5 mL of bacteriostatic saline) placed in the glabella (12 units), lateral canthi (8 units each), and medial forehead (12 units) on February 10, 2016. Two days later, on February 12, the patient had an intense pulsed light and full erbium:yttrium aluminum garnet (Erb:YAG) laser resurfacing treatment at a depth of 20 microns, and fractionated laser on 22% of the skin at a depth of 150 microns. The patient experienced significant swelling in the glabella, upper eyelid, and paranasal area from the procedure. In Figure 1, there is bilateral swelling, worse on the patient's left. The patient noticed on February 24 that her upper lip would not lift upward and called the office, concerned about this side effect. The patient was referred to a neurologist, however she declined the appointment. A neurologist was consulted by phone, and because he did not believe the patient to be at a risk for a stroke or other neurological issue, he recommended follow-up in a month if the symptoms did not start to resolve.

Figure 2A shows the patient's smile and her inability to lift her lip on both sides after the treatment, with the patient's left side lower and more affected than the right. The patient reported that this side effect started to improve approximately 1 month later, on March 21, with the ability to lift her upper lip to see her teeth, but she did not return to baseline for several months (Figure 2B).

DISCUSSION

Although BoNT treatment is an incredibly safe procedure, unwanted spread to adjacent muscles is a known but rare adverse effect. Internalization and breakage of the L chain disulfide bond in neuronal cells renders BoNT incapable of action elsewhere. Therefore, the spread of BoNT is most likely caused by unbound toxin in the extracellular space or by toxin within other cell types. For example, BoNT has been shown to cross, by transport vesicles, intact through epithelial cells without release to the cytosol.¹¹ In 1 study on mice that ingested BoNT-A, although BoNT was found in intestinal cells at 2 hours, there was still active toxin remaining in the epithelia after 24 hours.²⁸ Because BoNT is very stable, with a long half-life, in humans (reported to last up to 25 days in the serum of 1 patient), there is an opportunity for active BoNT not yet taken up by neuronal cells to spread to adjacent muscles under certain conditions.^12,13

The LLSAN requires very small doses to affect upper lip mobility, and is usually not intentionally injected, except in a minority of patients who request treatment for a gummy smile. The LLSAN is also near one of the most common treatment areas, the glabellar complex, which is composed of the procerus, corrugator and depressor supercilii. Therefore, it follows that upper lip ptosis would be a common clinically visible side effect from undesired BoNT spread. In fact, not only did it occur in this case, but it also was reported in 4 of the 5 case reports by Dal’Forno.²⁶

Another reason this region is at high risk for swelling-induced migration is the presence of the malar septum. Just as the malar septum holds blood tightly in this area after an injury above it (resulting in a “black eye”), its impermeable membrane also retains edema. Therefore, swelling in this area is much more severe than in the rest of the face. Although the LLSAN and the nasalis were not injected with BoNT in this case, the LLSAN originates close to the procerus and depressor supercilii; 1 recent publication shows an attachment of the LLSAN with either the procerus or depressor supercilii in a majority of cadaver specimens.²⁹ Given the proximity of these muscles, and the malar septum below keeping the edema and accompanying BoNT-A concentrated in this area, it is easy to see how this side effect could occur. Additionally, many areas of the face are partitioned as discrete sections, such as the superficial nasolabial and central forehead fat compartments, which may aid in the spread of molecules within these spaces.³⁰

Although there are other explanations for lip ptosis or muscle weakness, they are much less likely. Medical causes such as Bell's palsy or stroke would not have affected 1 muscle on both sides of the face. Migration to the LLSAN could happen with BoNT injection alone, if the injector aimed inferiorly and laterally in the procerus or inferiorly in the medial corrugator; however, the injector in this case always injects upward and laterally in the corrugator and in a straight perpendicular direction for the procerus, while pinching the muscle to prevent lateral spread. In addition, this patient had been injected previously with the same amount of BoNT A without any subsequent procedures by the same injector with the same injection technique, without any similar sequelae. In fact, the injector in this case has injected over 30,000 separate neuromodulator treatments in the glabella in over 5,000 individuals with the same injection pattern and technique without this type of complication. Finally, the degree of lip ptosis mimicked the degree of swelling, with more swelling on the patient's left causing more ptosis on that side, as shown in Figures 1, 2. It follows that more swelling would create a larger pathway for more neuromodulator to reach the LLSAN, similar to a larger reconstitution volume causing a larger spread of neuromodulator, as shown in the literature.

Last, it took approximately 40 days from injection for the LLSAN to start to regain mobility, with full movement of the lip occurring following a similar timeline to movement in the glabella, making the same cause for both more likely. In the report by Dal’Forno, time to resolution of reported adverse effects, including lip ptosis, took 40 to 75 days, similar to this case.

The injector had another case, in another location, with undesirable BoNT migration after a patient fell and hit the forehead 24 hours after neuromodulator injection in that area, causing a large bruise and swelling, with subsequent eyebrow ptosis on that side only. Similarly, a colleague consulted with the author regarding a patient who had a bee sting on the forehead 1 day after injection and also experienced a large amount of subsequent swelling, also resulting in brow ptosis on that side. Although either of these events may have occurred without the subsequent swelling, the 1-sided nature of the side effect lends plausibility to this novel hypothesis.

These reports are relevant to clinical practice, because there have been studies suggesting that Botox pretreatment with laser resurfacing yields improved results.²⁷ However, the optimal time before treatment for reducing potential migration risk has not been explored. In addition, intraoperative Botox for facial reconstructive surgery has been studied in a small retrospective study, and also discussed for use with Mohs surgeries for better scar outcomes. Although the few small studies done about resurfacing and reconstructive surgery (10 and 18 respectively) have reported treatment as safe, with no additional side effects, larger studies that specifically address high-risk areas for swelling have not been conducted to date.^27,31 If we can confirm this hypothesis, it would mean that a small amount of active BoNT that is not internalized persists in the area around the injection site for several days, while patients work out and go to sleep in different positions, all without any untoward effects unless a significant amount of swelling occurs; therefore many of the short-term postprocedure instructions are not necessary. Unfortunately, there are many different aftercare instructions for BoNT treatments, and most lack clear scientific evidence of their effectiveness or reduction of the severity or frequency of side effects. For instance, some physicians recommend patients keep their head elevated and refrain from exercise for 4 to 6 hours after the injection to prevent unwanted spread, although there is no evidence that changing position influences diffusion. In reality, if standing upside down could cause the BoNT to diffuse superiorly, then standing upright would cause the same gravity-dependent movement inferiorly, which is more likely to cause ptosis when the forehead is injected. One could then argue that lying on the back, something many physicians warn against, would be the least likely to cause gravity dependent migration of the BoNT inferiorly or superiorly. If the concern is that bending or exercising may cause facial swelling, the next question is about how much swelling may occur from these activities, especially shoe shopping, something many physicians regularly advise against, as an example of being careful not to bend downward. In 2004, 29% of the members of an expert consensus panel recommended that patients avoid bending their heads while 30% recommended avoiding exercise, despite the lack of evidence that this may cause unwanted side effects.³² Although physicians try to base their cosmetic practices on scientific evidence, unfortunately there is much about BoNT that we still do not know.

After this pateint case presented above, the author changed scheduling recommendations for BoNT pretreatment to at least 2 weeks before any procedure that may cause large amounts of swelling. In addition, the author provides no posttreatment instructions for BoNT patients regarding normal routine activities, because they would not cause the amount of swelling needed to assist in BoNT migration; the author only warns against procedures that are known to cause significant swelling, such as photodynamic therapy, laser resurfacing, intense pulsed light, microneedling, or deep chemical peels. Clearly there is more to learn, and further insights may lead to an ability to maximize clinical effects and decrease undesired side effects. There is a need to carefully examine what happens to toxin not taken up immediately and determine the exact timeline beyond injection at which even small amounts of BoNT persist, especially now that there is a new formulation (DaxibotulinumtoxinA) with a strong net positive charge, which is reported to remain localized extracellularly for even longer time periods.³³

Disclosures

Dr Siperstein is a consultant, clinical trial investigator, speaker, and trainer for Galderma (Lausanne, Switzerland) and Allergan (Irvine, CA).

Funding

The author received no financial support for the research, authorship, and publication of this article, including payment of the article processing charge.

REFERENCES

Author notes