Lip Hyaluronic Acid Fillers: A Meta-Analysis of Six-Month Response and Adverse Events

The authors concluded that HA injections were effective in increasing lip fullness through the six-month follow-up. A responder was someone whose score improved by at least one grade on a validated lip-fullness scale: the pooled estimates were 91% at 2 months and 74% at 6 months. At 12 months the estimate was 46% (95% CI 28–65%), with substantial uncertainty. GRADE certainty was very low for both effectiveness and adverse events. Common injection reactions and rare serious events were both reported; only 10 of 32 included records contributed to the effectiveness meta-analysis, and the search ended on December 31, 2018.

• The pooled response estimates were 91% at 2 months, 74% at 6 months, and 46% at 12 months. The 12-month 95% confidence interval was 28–65%; none of these estimates is an individual’s duration curve.
• In the abstract’s adverse-event summary, tenderness was reported in 88.8%, injection-site swelling in 74.3%, and bruising in 39.5%. Granulomatous foreign-body reactions, angioedema, and lip-necrosis case reports require separate attention.
• Of 32 included records, 10 studies with 1,228 participants supplied the effectiveness synthesis. Later evidence and current products fell outside the search cutoff.

Czumbel and colleagues examined HA fillers injected into the lips and perioral area. Their study tables name products, but the pooled findings do not establish a result for any brand or clinic.

Key findings

The authors searched the Cochrane Central Register of Controlled Trials (CENTRAL), Embase, and MEDLINE through December 31, 2018. Of 326 records identified, 259 remained after duplicate removal. The authors excluded 176 by title or abstract, assessed 83 full texts, and included 32 records. Ten studies—five randomized controlled trials and five cohorts—contributed 1,228 participants to the quantitative effectiveness synthesis. The other 22 records lacked suitable effect data or used measures that could not be combined.

A responder improved by at least one grade from baseline on a validated lip-fullness scale. These percentages estimate how many people met that threshold; they do not measure the percentage increase in lip volume. The estimate was 91% at 2 months (95% CI 85–96%), 74% at 6 months (66–82%), and 46% at 12 months (28–65%). Different studies contributed at each point. In a separate analysis of three no-treatment control studies, the 2-month estimate was 21% (6–40%). The authors viewed this as a possible placebo-effect signal in HA injection research.

The abstract reports adverse events among 1,496 participants: tenderness 88.8%, injection-site swelling 74.3%, and bruising 39.5%. The Results section uses 1,488 participants, with tenderness at 88.7%, contusion at 48.7%, and swelling again at 74.3%. The two summaries differ, so their figures cannot be combined. The review also reports rare serious reactions and rates certainty for both effectiveness and adverse events as very low.

What the six-month finding means

At the six-month follow-up, a majority in the pooled analysis still met the one-grade response threshold. The authors’ positive conclusion stops there. The 12-month estimate of 46% has a wide confidence interval and cannot establish a one-year result for an individual.

The 32 records differed in design, denominator, and follow-up. Case reports helped the authors examine rare and later adverse events; only 10 studies supplied compatible numbers for pooled fullness. The two sets of percentages do not describe one uniform study.

What did the review examine?

The pooled result came from a subset of records

The review sought studies of healthy adults older than 18 receiving HA filler in the lips or perioral area. Of the 32 included records, 10 effectiveness studies comprised five randomized controlled trials and five cohorts. Their mean ages ranged from 41 to 54 years, and follow-up from 12 to 48 weeks. The search covered English-language records. Two reviewers independently selected records and extracted data. That process does not change the final GRADE rating of very low.

The effectiveness studies excluded people with a history of allergy to injectable HA, previous semipermanent or permanent facial augmentation or aesthetic surgery, temporary facial filler or other facial cosmetic procedures within the preceding 24 months, or significant lip abnormalities. These are study exclusions, not universal eligibility rules.

Product names show what the studies used

The paper’s study tables name Juvéderm Ultra, Ultra XC, Ultra Plus, Volbella, and Volbella XC; Restylane, Restylane-L, and Restylane Lip Volume; Perlane; Belotero Balance Lidocaine and Belotero Intense Lidocaine; and Emervel. These are names of products appearing in included studies. The authors noted that the first two product families in this list appeared most often in the effectiveness studies. Some primary studies named product comparison groups, but this meta-analysis did not pool brand-specific effectiveness or safety comparisons. None of its 91%, 74%, or 46% estimates belongs to one of these products alone.

Response estimates across follow-up points

Five studies contributed to the 2-month estimate of 91% (95% CI 85–96%), and five to the 6-month estimate of 74% (66–82%). Eight studies supplied the separate 3-month estimate of 71% (55–87%). Different study sets produced each figure, so the move from 71% to 74% does not show a rebound in the same people.

Four studies supplied the 12-month estimate of 46% (28–65%). Some participants still met the threshold after a single injection, but the pooled estimate was lower than at 2 and 6 months and its interval was especially wide. Between-study heterogeneity (I²) was 82.7% at 2 months, 66.88% at 6 months, and 93.21% at 12 months. In three no-treatment control studies, the 2-month estimate was 21% (6–40%), with I² of 89.95%. The authors treated that response as a possible placebo-effect signal, not a reason to discard the injection-group findings.

Common reactions and rare serious events need separate descriptions

In the abstract’s summary of 1,496 participants, tenderness occurred in 88.8%, injection-site swelling in 74.3%, and bruising in 39.5%. The authors described common reactions—tenderness, swelling, bruising, injection-site masses, and pain—as mostly mild or moderate, generally resolving without treatment within weeks.

The Results section, by contrast, uses 1,488 participants. It lists 1,320 reports of tenderness (88.7%), 1,105 of swelling (74.3%), and 725 of contusion (48.7%), plus injection-site masses at 27.3% and pain at 19.7%. The denominator, tenderness value, and bruising-versus-contusion figures differ from the abstract. The paper does not explain why.

Rare reports included granulomatous foreign-body reactions (0.6%), herpes labialis (0.6%), tumor-like nodules (0.3%), and angioedema (0.3%). One study reported herpes labialis as its most frequent event at 17%; the authors could not explain why its figure was so much higher. They described granulomatous reactions as more serious and angioedema as potentially life-threatening. Three case reports described filler-associated lip necrosis, with no pooled rate. A low or absent pooled percentage does not rule out a risk. Nor does the course of common reactions apply to these rare events.

A positive conclusion with very low certainty

GRADE rated effectiveness evidence as very low. The synthesis included observational studies and substantial heterogeneity, with wide confidence intervals and suspected publication bias related to small-study effects. Adverse-event evidence also had very low certainty because of study design, risk of bias, and inconsistent reporting.

The authors lacked enough comparable subgroup data to examine injection volume, touch-up injections, technique, HA cross-linking, skin type, investigator experience, or assessment method comprehensively. The discussion describes more than eight HA products at five concentrations; the Results section refers to four injectable HA products. That internal difference cannot resolve product-level effects. Randomized trials were few, adverse-event definitions varied, and most prospective follow-up stopped within one year. Delayed or unreported events may have been missed. The authors called for higher-quality randomized trials and longer follow-up to clarify long-term effects.

The paper reports funding from Hungarian government and European Union academic programs, plus open-access publication support from the Central Library of Semmelweis University. The authors said the funders had no role in the research and declared no commercial or financial relationships that could be construed as a conflict of interest. These disclosures provide context; they endorse no product.

How to interpret the findings

1. Place in the literature

The authors described their work as the first meta-analysis to quantify HA lip-augmentation effectiveness. It brings the positive six-month finding and lower 12-month estimate together with common reactions and rare events from case reports.

2. Weight of the evidence

The review followed PRISMA reporting guidance and was registered in PROSPERO as CRD42018102899. Only 10 of 32 records contributed to pooled effectiveness, half were cohorts, heterogeneity was substantial, and both GRADE assessments were very low.

3. Meaning for an individual

Each responder proportion summarizes a different follow-up point, not an individual’s duration. Read it alongside the response definition, common reactions, and rare serious events rather than as a personal forecast.

4. What needs to be explained clearly

The authors’ positive conclusion covers the six-month follow-up. At twelve months, the estimate was 46% with a wide interval. “Mostly mild or moderate” describes common injection reactions; it does not describe angioedema or lip necrosis.

What the review supports

The pooled responder estimates were 91% at 2 months and 74% at 6 months. The authors concluded that HA injection increased lip fullness through the six-month follow-up. At 12 months, the estimate was 46%, with a wide interval and very-low-certainty evidence. None of these numbers, or a product name, determines an individual’s duration, adverse-event risk, or treatment choices.

Limits to keep in view

Only 10 of 32 included records supplied compatible effectiveness data; the other 22 had insufficient or noncomparable measures. Both GRADE assessments were very low. Observational designs, heterogeneity, wide intervals, suspected publication bias, risk of bias, and uneven adverse-event reporting limit inference. Too few suitable randomized trials and too little subgroup information were available to isolate technique, material, or patient factors. Different event definitions and unreported events may have led to undercounting. Most prospective follow-up ended within a year. The search stopped on December 31, 2018, leaving later evidence and the status of current products outside the review. The authors called for better randomized trials and longer observation.

References

1. Czumbel LM, et al. “Hyaluronic Acid Is an Effective Dermal Filler for Lip Augmentation: A Meta-Analysis.” Frontiers in Surgery. 2021;8:681028. PubMed · DOI.

Read the Korean original