Live·Open questions in longevity research
All news
TherapeuticsIndustry & policyClinical trials

FDA confirms that tumors in two patients are linked to Zolgensma gene therapy for spinal muscular atrophy: viral vector DNA integrated into the tumor genome at the GLI1 locus, while the prescribing information describes the risk as theoretical

9 November 2026· 261109009

FDA confirms that tumors in two patients are linked to Zolgensma gene therapy for spinal muscular atrophy: viral vector DNA integrated into the tumor genome at the GLI1 locus, while the prescribing information describes the risk as theoretical

On 6 November, FDA reported three tumors following Zolgensma treatment involving GLI1, a gene that can cause cancer when overactive. Tissue analysis of two of these tumors found viral vector DNA integrated into the chromosomes, with its promoter (a sequence that initiates gene transcription) positioned upstream of or within GLI1. FDA extended its assessment of tumor risk to Itvisma, which uses the same viral vector and is intended for people aged two years and older; no tumors have been reported following Itvisma treatment.

Zolgensma delivers a functional copy of SMN1 into cells inside an adeno-associated virus (AAV); this viral carrier is called a vector. More than 6000 patients have received the treatment, and its manufacturer, Novartis, has confirmed nine tumor reports since 2019. Most of the delivered DNA remains in the nucleus outside the chromosomes, and the prescribing information describes the risk of tumors from its integration as theoretical.

A study by Arul Chinnaiyan's group of another AAV therapy, FBX-101, described how integrated DNA activates GLI1. An MRI scan 26 months after treatment detected a tumor in a boy in which GLI1 expression exceeded that of every one of the 5143 tumors in the group's database. Targeted DNA sequencing found no cancer-driving mutations or gene fusions in the tumor, but long-read DNA sequencing explained the elevated expression: a vector fragment had integrated into the beginning of GLI1. The fragment contained a cytomegalovirus enhancer and a chicken beta-actin promoter (together known as CAG), which activate gene transcription, along with an RNA splice site, but lacked the therapeutic gene. Through this splice site, the cell joined the beginning of chicken beta-actin to GLI1, leading to excessive GLI1 expression. Pathologists have identified a similar fusion of human beta-actin with GLI1 in rare tumors unrelated to gene therapy.

On 28 January, Regenxbio reported that FDA had placed trials of its RGX-111 therapy on hold after a brain tumor developed in a boy who had received the treatment four years earlier. In the tumor, expression of PLAG1, a gene primarily active during embryonic development, was nearly 300 times higher than in the other nervous system tumors studied; a vector fragment had integrated into the gene.

Zolgensma, FBX-101, and RGX-111 carry the same pair of CAG elements. This pair had already raised concern: in a 2022 review, experts from the American Society of Gene and Cell Therapy noted that vectors containing CAG caused liver cancer in newborn mice, whereas a vector with a different promoter integrated into the same genomic region and did not cause tumors. The authors of the FBX-101 study note that both insertions identified in humans retained the CAG pair and the splice site.

The same authors suggest that the fragment lacking the therapeutic gene may have arisen during vector manufacturing. They propose removing the splice site from vector constructs and monitoring product purity: in a preprint on a liver biopsy from a child treated with Zolgensma, fragments of genes used to produce the vector amounted to 0,5–1% of the therapeutic gene copy count. Lindsey George, an author of the study on the tumor following RGX-111 treatment, and her colleagues consider it likely that tumor formation through AAV integration requires three conditions: a susceptible genomic site, features of the vector design, and a vulnerable biological state in the recipient.

FDA is assessing whether regulatory action is needed and whether the prescribing information and Novartis's long-term follow-up are adequate. It is also monitoring tumor risk in other similar AAV programs and continues to consider the benefits of both treatments to outweigh their risks for the approved indications. STAT describes a twenty-year debate over whether AAV can cause tumors: “That debate is over.”

Originally published on Telegram by Ukhvat NewsView on Telegram
Sources
#zolgensma#aav-vectors#gene-therapy#gli1#vector-integration#tumor-risk