ACI through three generations

If you are offered MACI today, you are being offered the third iteration of an idea from 1980s Gothenburg: harvest the patient's chondrocytes, expand them in culture, and re-implant them to regrow hyaline-like cartilage. What changed across generations is not the cells but the containment — and knowing the history protects you from marketing that borrows the wrong halo.

Generation 1: the periosteal flap (1987/1994)

Brittberg's 1994 New England Journal paper reported the first 23 patients: a cell suspension injected under a periosteal flap sutured watertight over the defect. Fourteen of 16 femoral condyle patients had good-to-excellent results and 11 of 15 femoral biopsies showed hyaline-like cartilage — but patellar results were poor (2 of 7). It was a first-in-human proof of concept, uncontrolled — not proof of superiority.

Long-term Swedish follow-up (224 respondents at 10–20 years) found 92% would have the surgery again and 74% felt the same or better, while Lysholm scores improved only modestly (60.3 to 69.5). Satisfaction endured; the scores were less impressive than the folklore — and it was a satisfaction survey with 34% non-response.

Why the flap died

A randomized trial of 68 patients settled it: clinical results with a periosteal cover were equivalent to a type I/III collagen membrane, but 36.4% of periosteal grafts needed re-operation for graft hypertrophy — versus zero with collagen. That finding, plus the morbidity of harvesting periosteum, removed generation 1 from practice. Generation 2 (ACI-C) was the same cell suspension under the collagen membrane instead.

Generation 3: MACI

MACI pre-seeds the cells onto the collagen membrane in the lab, and the loaded membrane is glued into the defect — no suturing a watertight cover, a shorter and simpler surgery. In the head-to-head RCT against ACI-C (91 patients), improvement at 1 year was comparable (Cincinnati +17.6 vs +19.6): MACI won on surgical practicality, not proven efficacy. The trials comparing generations were short and small — "MACI is better because it's newer" is convenience-backed, not outcome-backed.

The modern payoff came against microfracture: MACI won the SUMMIT RCT at 2 and 5 years in defects of 3 cm² and larger — though SUMMIT was sponsor-funded and open-label.

The null result that keeps the lineage honest

The Norwegian randomized trial followed generation-1 ACI versus microfracture to 14–15 years and found no difference — 17 failures versus 13 — with roughly half of survivors in both arms showing radiographic early osteoarthritis. Two lessons. First, cell therapy did not prevent arthritis. Second, marketing narratives about "30 years of ACI success" should not borrow the Brittberg 1994 halo for MACI claims: the generations differ, and the long-term randomized data for generation 1 are null.

Practical notes

Cell dose in classical ACI and MACI is on the order of 0.5–1 million cells per cm² (the Madrid group's 5 million/cm² variant has its own entry). US patients today get MACI, FDA-approved in 2016. Patellofemoral defects were generation 1's weak spot; modern series report better results there, but randomized patellofemoral evidence remains thin.

Still missing after 30 years: head-to-head trials of MACI against OATS or allograft, and any evidence that a cell therapy changes the long-term arthritis trajectory. The next generation — allogeneic or stem-cell-derived chondrocytes to remove the two-stage burden — is in early-phase trials, with nothing at approval level yet.