CaHA: the filling evidence holds up; the 'collagen regeneration' story came out the other way

The 2008 trial enrolled 60 patients at two European clinics, injected CaHA on one side of the face and non-animal stabilised hyaluronic acid (NASHA) on the other, and had blinded evaluators assess at 6, 9 and 12 months.

CaHA came out ahead at every time point. At 12 months, folds still improved or better: 79% for CaHA against 43% for NASHA (p < .0001), using 30% less total volume. Evaluators judged CaHA superior in 47% of patients and inferior in 5%.

The 2025 trial randomised 210 subjects in a multicentre, double-blind study in China. Syringes were made identical and labels coded, so neither the injecting physician nor the patient knew which product was which. 188 completed 12 months of follow-up.

At 24 weeks the wrinkle-scale improvement rates were 84.04% against 78.72% by investigators and 72.34% against 70.21% by an independent review committee, and non-inferiority was met (P > .05).

Two cautions when reading them together. First, the control is always another filler — no trial compares against doing nothing. Second, the 2025 trial used a different company's CaHA product, not Radiesse. Sharing a material does not make two products the same.

A 2024 systematic review of CaHA sorted the evidence by facial region. It concluded that CaHA may be considered safe and effective for cheeks, jawline, HIV-related facial lipoatrophy and nasolabial folds.

And then this, as the last line of the abstract: despite the recent trend, guidelines and safety profile of diluted and hyperdiluted Radiesse, no randomised controlled trials have been published.

That sentence matters. What is marketed as a 'Radiesse skin booster' or 'collagen booster' is generally this diluted use. But the two randomised trials above were done with undiluted product, filling folds.

This is where evidence gets carried across. Results from filler trials are quoted as if they supported the diluted technique. Same material, but a different concentration, a different tissue plane and a different goal. It is exactly why this site puts concentration and vehicle ahead of the name of a substance.

The 2025 study is a 13-week, randomised, single-centre comparative trial. Poly-L-lactic acid (PLLA) and CaHA were injected into nasolabial folds, and punch biopsies at baseline and day 90 were compared for gene expression. Participants were selected to have matching wrinkle grades on both sides.

The two materials did not take the same path.

• PLLA stimulated more extracellular matrix components with less inflammatory response. The authors read this as a regenerative pathway.
• CaHA elicited a stronger inflammatory response, which the authors note could diminish tissue regeneration. Their conclusion is explicit: CaHA did not show evidence of tissue regeneration and upregulated more genes in pro-inflammatory pathways.

That cuts directly against the 'it regenerates collagen' account. It is tissue-level evidence from biopsy, which is not light.

Two caveats for reading it. This is a head-to-head between two products, and we could not identify the funding source from the abstract; a comparative trial that favours one product should be read alongside the question of who ran it. And gene expression is not the same thing as a clinical result — as the first item shows, that CaHA fills a fold is established separately.

What is established is 'it fills'. What is not established is 'it regenerates'. That is why these sit as two items.

A 2026 systematic review collected reports of visual loss from vascular occlusion with non-hyaluronic-acid injectables.

The counts as published:

• CaHA: at least 11 published cases of vascular occlusion causing visual impairment, most involving the nasal dorsum.
• PLLA: two confirmed reports.
• PDLLA-CMC: two. PDLLA-HA: one.
• PCL: no published cases of blindness, one case of facial artery embolism.

That CaHA tops this list should be written down as it is. Case counts follow usage as well as risk, so they cannot be read straight off as a risk ranking — but they do tell you that this happens, and where the injection was when it did.

There is also a decisive difference from hyaluronic acid. HA can be dissolved with hyaluronidase; CaHA has no such antidote. What you can do about an occlusion is not the same.

The authors set out a management protocol: sound knowledge of anatomy and product characteristics, prompt recognition of symptoms, immediate intraocular pressure-lowering agents and referral to an ophthalmologist, plus hyperbaric oxygen, anti-inflammatory and anticoagulation support.