Is there a risk of cancer from the therapy itself?
The most feared long-term risk of gene-modified T-cell therapies is that the viral vector used to deliver the new receptor might accidentally insert into a cancer-causing gene and trigger a secondary malignancy. The largest and most comprehensive study here directly addressed this: across 783 patients from 38 trials, only 18 (2.3%) developed a secondary cancer, and in every case where the tumor could be tested, there was no evidence that the vector had caused it [1]. One T-cell lymphoma was found, but the malignant cells did not contain the vector [1]. This strongly suggests that the risk of the therapy itself causing cancer is very low.
A separate study with 43 patients followed for up to 9 years reported that long-term adverse events were rare, with the main lasting effect being B-cell depletion and low antibody levels (hypogammaglobulinemia) [4]. Together, these two studies—the largest of their kind—converge on the same conclusion: the risk of a secondary malignancy directly caused by the gene therapy is minimal.
What about long-term effects on the brain and mental health?
Here the picture is more mixed. One study of 40 long-term survivors (1–5 years after treatment) found that while average scores for mental health, anxiety, and depression were normal, nearly half (47.5%) reported at least one clinically meaningful problem: cognitive difficulty, depression, or anxiety [2]. Younger age and pre-existing anxiety or depression were risk factors [2]. This suggests that a significant minority of patients do struggle with lasting neuropsychiatric effects.
However, a different study of 19 patients who were disease-free 2 years after treatment found no new neurologic deficits, no changes on brain MRI, and no decline in cognitive test scores—even among the 11 who had experienced acute neurotoxicity right after treatment [3]. Anxiety scores actually improved over time [3]. The apparent conflict between these two studies likely reflects differences in measurement: the first used patient-reported surveys (which capture subjective experience), while the second used objective neurological exams and cognitive tests. Both are valid, but they measure different things. The takeaway: objective brain function appears stable, but many patients still feel cognitively or emotionally worse.
Can we predict these risks before human trials?
Current preclinical models have historically failed to predict the severe toxicities seen in humans, such as cytokine release syndrome and neurotoxicity [5]. A consortium of experts (T2EVOLVE) has called for better models, including humanized mice, to improve prediction of both safety and efficacy [5]. This means that some risks may only become clear in human trials, which is why long-term follow-up studies like those above are so important.
On the positive side, newer genetic engineering approaches are being developed to reduce risks. One platform uses CRISPR to remove both endogenous T-cell receptors, which prevents the engineered cells from pairing with natural receptors and causing off-target autoimmune reactions [6]. This approach enhanced tumor clearance in mice and prevented graft-versus-host disease, and genomic analysis showed no dangerous changes to the vector integration sites [6]. While still preclinical, this points to a future where TCR therapies may be inherently safer.
About These Sources
This answer is built on 6 peer-reviewed studies — published from 2021 to 2026, 2 from 2024 or later, 3 in Q1 journals, collectively cited 133 times — selected as the most relevant from 6 studies that passed quality screening, drawn from 60 papers retrieved from a database of over 500 million.
Sources used in this answer
Long-term safety of lentiviral or gammaretroviral gene-modified T cell therapies
In the largest study here (783 patients, 2,200+ patient-years), only 2.3% developed secondary malignancies, and none were linked to the gene therapy vector; one T-cell lymphoma was found but lacked the vector.
Patient-Reported Neuropsychiatric Outcomes of Long-Term Survivors after Chimeric Antigen Receptor T Cell Therapy.
Among 40 long-term survivors (1–5 years), average mental health scores were normal, but 47.5% reported at least one clinically meaningful cognitive difficulty, depression, or anxiety; younger age and pre-existing anxiety/depression were risk factors.
Long-term Neurologic Safety in Patients With B-Cell Lymphoma Treated With Anti-CD19 Chimeric Antigen Receptor T-Cell Therapy
In 19 disease-free patients at 2 years, no new neurologic deficits, MRI changes, or cognitive decline were found, even in those who had acute neurotoxicity; anxiety scores improved.
Long-Term Follow-Up of Anti-CD19 Chimeric Antigen Receptor T-Cell Therapy.
In 43 patients followed up to 9 years, 51% of treatments led to remissions lasting >3 years, and late adverse events were rare except for B-cell depletion and hypogammaglobulinemia.
Time to evolve: predicting engineered T cell-associated toxicity with next-generation models
Existing preclinical models have failed to predict severe toxicities like cytokine release syndrome and neurotoxicity; the T2EVOLVE consortium advocates for improved humanized mouse models.
A universal platform for simultaneous TCRα/β removal enables safer and more potent TCR therapies and autoimmune modeling
A CRISPR platform that removes both endogenous TCR chains achieved >90% deletion efficiency, enhanced tumor clearance in mice, prevented GVHD, and showed no dangerous changes to vector integration sites.