"We were still holding commercialization meetings until just two days before the announcement.""It is statistically rare for this level of effect to appear in the placebo group, and it was a result that was hard to accept under common clinical sense."Jeon Seung-ho, co-CEO of Kolon TissueGene
The osteoarthritis cell and gene therapy "TG-C" from Kolon TissueGene(950160) failed to achieve statistical significance versus placebo in one of two phase 3 trials in the United States. The company held a press briefing on the 21st and called it "half a success, neither a success nor a failure," but the market's view is cold.
As of 10:34 a.m. that day, Kolon TissueGene was trading at 42,900 won, down 29.90% (18,300 won) from the previous trading day.
The seven-year drama of a near-miraculous comeback that began with the "Invossa incident" in 2019 has now entered a phase where everything rides on the remaining single clinical trial and on negotiating power with the U.S. Food and Drug Administration (FDA).
◇ "The drug effect held, but"… phase 3 undone by a surge in placebo effect
Kolon TissueGene announced topline (key endpoint) results of the U.S. phase 3 trial (TGC15302; 531 patients) the previous day via a regulatory filing. At 12 months, neither the pain index (VAS), a co-primary efficacy endpoint, nor the WOMAC total score, which comprehensively evaluates pain, stiffness, and physical function, achieved statistical significance versus placebo.
The magnitude of pain improvement in the TG-C arm itself was not much different from phase 2. On the VAS, pain reduction was 38.7 points in this phase 3 versus 39.7 points in phase 2, maintaining a similar level. But the placebo group's pain reduction jumped from 24.3 points in phase 2 to 39.2 points in this phase 3, effectively erasing the between-arm difference.
TG-C's relative effect, which outperformed placebo by 15.4 points in phase 2, flipped to -0.5 point in this phase 3.
Jeon Seung-ho, co-CEO of Kolon TissueGene, said at the press briefing that day, "I think it is half a success, neither a success nor a failure," and added, "We will closely analyze the causes of the larger-than-expected placebo response and decide our future approval strategy and commercialization direction."
He went on, "The approval timeline may be somewhat delayed from the original plan, but our push for commercialization is unchanged." Kolon TissueGene had initially planned to submit a biologics license application (BLA) to the FDA in the first quarter of 2027.
The industry views it as not uncommon for placebo responses to be larger than expected in large-scale osteoarthritis trials. However, cases where the placebo group's improvement expands to a level similar to the treatment arm, as in this case, are considered unusual. While heterogeneity of patient populations or lifestyle changes during enrollment across multiple clinical sites could have influenced the results, there is a view that such explanations alone will be hard to persuade the FDA.
No new safety signals were identified. The rate of total knee arthroplasty was 0.6% in the TG-C arm versus 5.3% in the placebo arm, lower in the TG-C arm. The company interpreted this as suggesting the potential to delay joint replacement surgery, but the prevailing view in the industry is that it is insufficient to overturn the failure to meet the primary efficacy endpoints.
◇ Fate hangs on the second trial… a test of FDA negotiating power
What remains now is a single U.S. phase 3 trial, TGC12301 (535 patients). The topline results of this study are expected to be a watershed for Kolon TissueGene's approval strategy.
If TGC12301 achieves statistical significance versus placebo for both the pain index (VAS) and the WOMAC total score, Kolon TissueGene will be able to discuss an approval strategy with the FDA based on the outcome of "one success and one failure." However, it is uncertain whether a BLA would be accepted on the strength of a single success. Identifying the reasons for the placebo-response differences between the two trials, additional analyses, submission of supplementary materials, or the need for additional trials are expected to be central points of discussion.
Conversely, if TGC12301 also fails to demonstrate efficacy, both confirmatory trials would fail to prove improvement in pain and function. In that case, the assessment is that FDA approval will be difficult to expect under the current development strategy alone, and the company will likely face the prospect of having to redraw its overall commercialization strategy.
Jeon said, "The guideline the FDA recently released includes language that approval could be considered even from a single clinical trial if sufficiently robust results are obtained," adding, "But given the significant uncertainty, even if positive results come from the second trial, the company plans to take a conservative approach." He left open the possibility of additional trials.
Attention is also focusing on the undisclosed 24-month follow-up results. In particular, the key is how much the structural improvement indicators (JSW, MRI) have improved.
For Kolon TissueGene to be recognized as the "world's first DMOAD (a disease-modifying osteoarthritis drug that slows disease progression)," it must demonstrate not only pain relief but also structural improvements such as cartilage regeneration. Given that TGC15302 failed to meet the primary endpoints of pain and function improvement, some analysts say that even if structural endpoints are positive, it will not immediately translate into DMOAD approval prospects.
In phase 2, structural improvement indicators such as reduced progression of cartilage damage on MRI and suppression of synovitis did not achieve statistical significance. However, because this phase 3 enrolled more than 10 times as many patients—rising from 102 to 1,066—the company expects the statistical reliability of the structural endpoints to improve over the previous trial.
As the approval timeline looks more likely to be delayed, the manufacturing contract with the global contract manufacturing organization (CMO) Lonza has also emerged as a variable. If the BLA filing schedule is pushed back, the production schedule and contract terms may need to be renegotiated.
Noh Moon-jong, co-CEO of Kolon TissueGene, told ChosunBiz in response to related inquiries, "We are currently preparing a detailed analysis of the 24-month follow-up data," and added, "We also plan to discuss the contract with Lonza in line with the future approval timeline."
◇ Non-clinical variables pile up… CRO controversy and Invossa lawsuits
Separate from the clinical outcomes, the method of data analysis also stirred controversy in this announcement. Kolon TissueGene said its in-house analysis team conducted the topline analysis instead of outsourcing it to a conventional contract research organization (CRO) or an external statistics firm.
If a company has in-house biostatistics staff, it can perform internal analyses after database lock. However, with stock volatility expanding sharply ahead of the topline release, news of the internal analysis approach heightened market skepticism. Kolon TissueGene shares hit a 52-week high of 138,800 won on May 12, then fell to as low as 55,400 won intraday on the 16th, approaching the lower limit.
Some in the market say it is hard to rule out the possibility that financial authorities will look into the circumstances. In particular, when announcing results of the second trial (TGC12301), demands will likely grow to disclose the analyzing party and verification procedures more transparently.
The company dismissed the suspicions. Noh said, "We simply conducted an in-house analysis because we have biostatistics staff inside; there was no special reason," adding, "While the overall analysis was performed internally, some statistical items were rechecked through external vendors."
He added, "In the forthcoming causal analysis (related to the placebo group), we also plan to review the entire dataset again." Kim Jeong-in, chief financial officer (CFO) of Kolon TissueGene, said, "A small number of internal staff performed the statistical analysis, and we restricted related parties' stock transactions, so we judge there was no possibility of internal information leakage."
On top of that, legal risks surrounding the past Invossa incident are resurfacing. On the 9th, the Civil Division 29 of the Seoul Central District Court ruled entirely in favor of 139 patients who received Invossa in their damages lawsuit against Kolon Life Science and Kolon TissueGene.
It was a first-instance ruling handed down about seven years after the lawsuit was filed in Aug. 2019. The court recognized a manufacturing defect in that, although the label at the time of approval stated "cartilage-derived cells," the product was actually manufactured with kidney-derived cells. Kolon Life Science said it would review whether to appeal after receiving the written judgment.
It remains uncertain whether this ruling will affect the remaining two damages suits related to Invossa. If the cases proceed to appeals and the Supreme Court, the legal and financial burdens could be prolonged, and observers say the ruling could influence subsequent cases to some extent.
However, some in the industry argue that "the U.S. phase 3 trials disclosed to participants in advance that the cell origin was kidney-derived and proceeded after regulatory review, so it is hard to compare them on the same level as the Invossa incident in Korea in 2017."