In 1514, Albrecht Dürer drew his mother, Barbara, in charcoal at the age of 63. He did not soften the gauntness of an aging face, and the hollowing at the side of the forehead is part of what the drawing shows. It is a look that every era has recognized in its own way.

People who ask me about temple filler describe the same thing in modern terms. In photographs, the sides of the forehead look sunken and the bone seems to show. The literature calls this temple hollowing: a loss of volume in the temporalis muscle and the temporal fat pad that leaves bony margins visible and a gaunt, "skeletonized" appearance (Montes et al., 2025).

My argument is that temple filler works for temple hollowing, and the trial evidence is clear on that point. What the studies have not settled is which tissue layer the filler belongs in, and the safety evidence is thinner than the efficacy evidence. Three recent papers carry the case: one randomized controlled trial and two 2026 meta-analyses.

Does temple filler work for temple hollowing? What the randomized trial found

Yes. In a randomized controlled trial, 80.4% of treated participants improved by at least one grade at month 3, compared with 13.5% of untreated controls (Montes et al., 2025). The trial enrolled 171 adults at 15 sites in the United States and randomized them 2:1 to a hyaluronic acid filler called VYC-20L (n = 113) or to no treatment (n = 58) (Montes et al., 2025). Blinded evaluators scored each temple on the Allergan Temple Hollowing Scale, a validated 5-point scale, and a responder was someone who improved by at least one grade in both temples (Montes et al., 2025). The difference between groups was statistically significant (P < 0.0001) (Montes et al., 2025).

The other measures pointed the same way. At month 3, the blinded evaluators rated 83.8% of treated participants as improved or much improved on the Global Aesthetic Improvement Scale, against 10.7% of controls (Montes et al., 2025). Participants gave themselves the same rating in 92.9% of the treated group and 2.0% of the control group (Montes et al., 2025). On the FACE-Q questionnaire for satisfaction with the temples, the treated group's score rose by a mean of 42.7 points (95% CI, 37.4–48.0), while the control group's fell by 4.2 points (Montes et al., 2025).

Because the control group received no injection, the evaluators could be blinded and the participants could not. That is a limit of this kind of design, and it is worth remembering when reading the participants' own ratings.

How long does temple filler last? Thirteen months of follow-up

In the same trial, 73.3% of treated participants were still responders on the temple hollowing scale at month 13 (Montes et al., 2025). The responder rate was 81.7% at month 1 against 23.6% in the untreated group, and the treated group's effect was visible from that first month and lasted more than a year (Montes et al., 2025). At month 13, 72.2% were rated improved or much improved by the blinded evaluators and 78.9% by the participants (Montes et al., 2025).

The treatment itself is worth describing, because it shapes how to read these numbers. The filler was injected as boluses (small deposits) into the supraperiosteal plane, directly on the bone, with an optional touch-up at day 30 that 65.2% of treated participants received (Montes et al., 2025). The median total volume for both temples was 3.65 mL for the initial treatment and touch-up combined (Montes et al., 2025). Participants who had repeated treatment later needed approximately half the injection volume, according to the authors (Montes et al., 2025).

One limit applies here. The untreated controls were followed for only 3 months, after which they could choose treatment, so everything after month 3 is a comparison with baseline, not with controls (Montes et al., 2025).

What do pooled studies say about temple filler? A 14-study meta-analysis

A 2026 meta-analysis of 14 studies reported a pooled responder rate of 86.3% on the Global Aesthetic Improvement Scale (95% CI, 82.5–90.1) (Hashemloo & Milanifard, 2026b). The studies included 3 randomized trials, 9 prospective studies and 2 retrospective studies, covering more than 500 patients treated with hyaluronic acid or calcium hydroxylapatite fillers (Hashemloo & Milanifard, 2026b). Hyaluronic acid fillers had a pooled rate of 86.1% and calcium hydroxylapatite, based on only two studies, 84.1%, and the difference between filler types was not statistically significant (P = 0.687) (Hashemloo & Milanifard, 2026b).

Two cautions come with that number. A responder on this scale is anyone rated "improved" or better, which is a low bar, and heterogeneity (how much the studies disagree with one another) was substantial (I² = 70.06%) (Hashemloo & Milanifard, 2026b). The authors also note that responder definitions varied, with some studies using physician ratings, some patient ratings, and some both (Hashemloo & Milanifard, 2026b).

The review reports that injection volume was significantly associated with effectiveness in meta-regression. The coefficient, however, was negative (−0.004; P = 0.019), so this is not evidence that more filler gives a better result (Hashemloo & Milanifard, 2026b). The authors attribute much of the variation in volume between studies to differences in the severity of hollowing, the product, the technique and the aesthetic goals (Hashemloo & Milanifard, 2026b).

Which layer is right for temple filler? The studies do not agree

The randomized trial placed filler deep, on the bone. A second 2026 meta-analysis looked at a shallower plane, between the superficial and deep temporal fascia, and reported 85% efficacy (95% CI, 81–91%) (Hashemloo & Milanifard, 2026a). That review pooled 7 studies with 109 participants in total, and four of the seven were reports of a single patient (Hashemloo & Milanifard, 2026a). Subgroup results looked favorable for 1.5 mL per side (94%, from one study) and for an 18-gauge cannula (94%, from one study), but the tests for differences between subgroups were not significant (P = 0.15 for volume and P = 0.32 for cannula size) (Hashemloo & Milanifard, 2026a).

The authors argue that the interfascial plane lets a smaller amount of filler do the work and may be safer, and they note that the 21-gauge cannula, used in five of the seven studies, had an efficacy of 86% (Hashemloo & Milanifard, 2026a). The trial's authors give a different reason for their choice. They write that a deep supraperiosteal injection may reduce the risk of intravascular injection, since the vessels in the temple lie more superficially, while accepting that deeper placement may need more product (Montes et al., 2025).

The other 2026 review adds a further layer to the debate. It reports that most papers describe injection between the temporalis muscle and the bone, but that recent anatomical evidence suggests this space may not be a true potential plane, and that filler may often end up inside the muscle (Hashemloo & Milanifard, 2026b). It adds that deep injections are now regarded as less favorable by some, and that injection planes were not consistently reported in the pooled studies (Hashemloo & Milanifard, 2026b).

No head-to-head trial compares these layers. What we have is a well-run trial of the deep approach and a small, mostly case-level body of work on the shallow one.

Is temple filler safe? What the trial and the reviews report

In the randomized trial, adverse events were mostly mild, and no vascular occlusion, serious treatment-related event, or late-onset event was reported (Montes et al., 2025). Among treated participants who kept an e-diary, 59.0% (95 of 161) recorded at least one injection-site response after the initial treatment, most commonly pain on injection (50.9%), tenderness (49.7%) and redness (41.0%), and most were mild or moderate and resolved within a week (Montes et al., 2025).

Treatment-related adverse events occurred in 29 participants (17.6%), most often jaw pain (6.1%) and headache (4.8%) (Montes et al., 2025). Four participants had mild treatment-related events that lasted longer than 30 days, including mild trismus (reduced jaw opening) and an injection-site mass, and all resolved without intervention (Montes et al., 2025). Vision tests showed no clinically meaningful change, and jaw function scores did not change significantly (Montes et al., 2025). The authors explain the jaw symptoms by the position of the temporalis and masseter muscles, which sit in and around this region (Montes et al., 2025).

The pooled reviews are less reassuring about how much we know. The 2026 meta-analysis of 14 studies reports mild and temporary complications such as edema, bruising, tenderness, erythema and swelling, and found no consistent report of severe events such as skin necrosis or vision loss (Hashemloo & Milanifard, 2026b). Its authors still say that safety reporting was limited and heterogeneous, and that conclusions about overall safety should be interpreted with caution (Hashemloo & Milanifard, 2026b).

Anatomy explains the caution. The trial's authors list the superficial temporal artery, the middle temporal vein and the deep temporal arteries as structures to protect, note substantial variation in facial arterial anatomy between individuals, and recommend palpation and visual inspection of the superficial temporal artery before injecting (Montes et al., 2025). The interfascial review cites literature that puts the internal diameter of this artery, especially its anterior branch, at less than 1 mm, and it considers ultrasound imaging important (Hashemloo & Milanifard, 2026a). Its pooled result concerns efficacy. The comments on artery diameter come from the cited anatomical literature, not from a pooled measurement.

The strongest objection

Some will say that a randomized trial with a 13-month follow-up and a meta-analysis of 14 studies close the question. That is partly true. Temple filler has now been tested against no treatment under evaluator blinding, and the benefit on the scale used was large and lasting (Montes et al., 2025).

But each source has limits. The trial was funded by the manufacturer, most of its authors report investigator or consultant roles with the company, and two are company employees (Montes et al., 2025). Participants were mostly White (83.0%) and women (83.6%), the authors note the lack of two-dimensional photographs, and the control group was observed for 3 months only (Montes et al., 2025). The meta-analyses carry substantial heterogeneity, and the three randomized trials in the larger one were judged to have "some concerns" about deviations from the intended intervention (Hashemloo & Milanifard, 2026b). One of those three trials is a 2024 conference abstract of a 13-month, multicenter, evaluator-blinded pivotal study that appears to be the earlier report of the same trial discussed above, so the two sources are not fully independent (Hashemloo & Milanifard, 2026b; Montes et al., 2025).

And the safety evidence is mostly an absence of reports. That is not the same as a measured rate.

What should you ask before temple filler?

A good consultation about temple filler starts with the question of which layer is proposed and why. The honest answer, given the literature, is that the deep supraperiosteal plane has the better trial and the interfascial plane has an argument about volume and safety, and that the two have not been compared directly (Hashemloo & Milanifard, 2026a; Montes et al., 2025).

It is also reasonable to ask which product will be used, since the randomized trial tested a single product and its results should not be assumed for others (Montes et al., 2025). Ask how the injector handles the temple's vessels, including whether the superficial temporal artery is palpated or inspected first (Montes et al., 2025), and whether ultrasound is used, which one review considers important (Hashemloo & Milanifard, 2026a). Hyaluronic acid can be reversed with hyaluronidase, so it is fair to ask whether it is available in the room (Montes et al., 2025).

Finally, ask what a normal recovery looks like. In the trial, most injection-site reactions settled within a week, and temporary jaw discomfort and headache were the most common treatment-related events (Montes et al., 2025).

The Clinical Bottom Line

Temple filler is a reasonable, evidence-supported option for people with temple hollowing. In the randomized trial, most treated participants improved and most still had a response at 13 months (Montes et al., 2025), and the pooled data from 14 studies point the same way (Hashemloo & Milanifard, 2026b).

The caution is about what the studies leave open. The layer of injection is unsettled, the safety evidence rests on a single industry-funded trial and on reviews that describe their own safety data as limited, and some of the numbers behind the pooled results are small (Hashemloo & Milanifard, 2026a; Hashemloo & Milanifard, 2026b; Montes et al., 2025). What we still lack are direct comparisons of injection planes, independent trials, and a measured rate for the rare vascular events that matter most.

Dürer drew what he saw and did not flatter it. Temple filler should be judged the same way: by what it restores, and by how plainly its evidence is stated.

For clinical assessments and consultation in Seoul: itsdrbock.com

References

Hashemloo, A., & Milanifard, M. (2026a). Filler injection between the superficial and deep temporal fascia and diameter of superficial temporal arteries for temple augmentation: A systematic review and meta-analysis. Journal of Cosmetic Dermatology, 25(1), Article e70632. https://doi.org/10.1111/jocd.70632

Hashemloo, A., & Milanifard, M. (2026b). Temporal filler injection for lifting purposes: A systematic review and meta-analysis. Journal of Cosmetic Dermatology, 25(7), Article e71072. https://doi.org/10.1111/jocd.71072

Montes, J. R., Hooper, D., Jones, D., Biesman, B., Baumann, L., Beer, K., Yoelin, S., Leung, K., & Chawla, S. (2025). Improvement in temple hollowing with VYC-20L hyaluronic acid filler: A multicenter randomized controlled trial of safety and effectiveness. Plastic and Reconstructive Surgery, 156(1), 25e–36e. https://doi.org/10.1097/PRS.0000000000011957