After 2 Deaths and Reports of Brain Tumors, How Dangerous Is Gene Therapy, Really?

Many serious genetic diseases happen when both copies of a gene, one from each parent, are mutated, so there’s no working backup copy left to make the protein we need. These diseases are individually rare, so there’s not much commercial incentive to develop treatments for them, and gene therapy is often the only option. Sometimes it’s the only thing that lets a child survive to adulthood.

Gene therapy delivers a healthy copy of the gene into cells so they can make the missing protein themselves. The main bottleneck is delivery efficiency, and the best tool we have for that is Adeno-Associated Virus, or AAV. For treatment, we use a stripped-down version of the virus so it can no longer replicate, so it can’t spread or cause disease, but it’s still efficient at getting genetic cargo into cells and keeping it active for years. That combination of safety, efficiency, and durability has made AAV the backbone of the gene therapy field, with hundreds of clinical trials and thousands of patients treated over the past decade.

But as more trials happen, rare adverse events are accumulating and getting attention. There are two main ways this happens. First, the immune system can recognize AAV as foreign and, in rare cases, mount a strong enough response to damage healthy tissue, which is dangerous if that tissue is the heart, lungs, liver, or kidneys. Second, in less than 1% of cases, a cell mistakenly inserts the viral DNA into its own genome during routine repair. Most of the genome isn’t sensitive to this, but on rare occasions the insertion switches on a cancer-causing gene and leads to tumor formation. Both risks go up with higher viral doses, which is why figuring out the right dose, frequency, and tissue target is central to clinical trial design and why the first phase of any trial is entirely focused on safety before efficacy is ever tested.

That safety-first process is the right one, but it’s also slow, often taking years. For patients with fatal, fast-progressing diseases, that timeline itself becomes a risk, and some lives have been lost to the wait, not the therapy. This is a genuinely hard tradeoff for policymakers: more safety checks can mean missing the window to save someone, but fewer checks can mean real danger too. Rather than looser regulation, what might help is faster, well-resourced external review for these high-stakes cases.

Meanwhile, a lot of research has gone into alternatives. Small RNAs like siRNA and antisense oligonucleotides can’t integrate into DNA, so they avoid that risk, and both are already FDA approved for certain conditions, though they mainly reduce gene activity rather than replace a missing gene outright. mRNA and self-amplifying RNA can deliver a whole gene the way AAV does, but current versions are less stable or don’t last as long. None fully match AAV’s combination of safety, efficiency, and longevity yet, but given the pace of progress in RNA medicine, real alternatives are probably coming. For now, AAV is still the most powerful tool we have, and it’s only getting safer.

The vast majority of genetic diseases result from disruption of a single gene, yet their consequences can be devastating and lead to premature death. Despite the severity of these conditions, therapeutic options for most rare genetic diseases remain extremely limited or entirely absent. Gene therapy has changed this trajectory and is already delivering life-changing, and in some cases potentially curative, treatments for diseases that were previously considered untreatable.

Gene therapies are not inherently more dangerous than other treatments, but they present unique risks because they are designed to produce long-lasting, and often irreversible, biological changes. The risk also varies considerably depending on the therapeutic platform, the gene being targeted, the delivery system, the dose, and the underlying disease. Importantly, the risks of treatment must always be weighed against the risks of the disease itself. For a child with a rapidly progressive and ultimately fatal genetic disorder, delaying treatment can itself have irreversible consequences.

Alongside remarkable successes, the gene therapy field has faced significant setbacks. Adverse events understandably receive substantial attention because of the novelty and potential risks of these technologies. However, disproportionate attention to individual events can also create periods of hesitation that slow scientific and regulatory progress for patients who have few or no alternatives. Importantly, adverse events have provided critical lessons that have strengthened the field. Past complications have driven technological innovation, improved our understanding of gene delivery and immune responses, and led to increasingly rigorous regulatory oversight and clinical monitoring.

The recently reported deaths of children receiving gene therapies in China are therefore an important reminder that rapid development cannot come at the expense of appropriate safety evaluation and careful patient monitoring. Serious adverse events could occur anywhere, including in the United States, because no medical intervention is without risk. However, rigorous preclinical testing, regulatory review, carefully designed clinical trials, independent safety oversight, appropriate patient selection, and long-term follow-up are essential safeguards designed to minimize those risks and identify complications as early as possible.

At the same time, requiring every new gene therapy to repeat extensive testing when substantial safety knowledge already exists for a well-characterized platform can unnecessarily delay access. The appropriate path forward is a risk-based development paradigm that leverages prior knowledge while maintaining targeted, product-specific safety assessments. Our responsibility is not to eliminate all risk, which is impossible, but to ensure that risks are scientifically justified, carefully monitored, and proportionate to the severity and urgency of the disease. For children facing otherwise incurable and rapidly progressive disorders, both excessive risk and excessive delay can have profound consequences.

Gene therapy has made remarkable strides in improving the prognosis for children with rare diseases that, until recently, had few or no effective treatment options. We have made enormous progress in safety, oversight, and trial design since the tragic death of Jesse Gelsinger in 1999, but these are still novel therapeutics, and there are risks—both short- and long-term—that we may not yet fully understand.

For some families, it is entirely reasonable to accept a degree of uncertainty when the potential benefit is the opportunity to substantially alter the course of a severe childhood disease. For other families, the possibility of causing an unforeseen or irreversible downstream harm may feel unacceptable. I don’t think there is one right answer. Families have different values, different circumstances, and different tolerances for uncertainty, particularly when making decisions on behalf of a child.

That is why I believe ongoing informed consent and, when appropriate, assent need to be at the center of gene therapy trials. Consent for a potentially high-risk, first-in-human or otherwise experimental therapy cannot be a single event in which a family signs a document. It should be an ongoing conversation. Families need information presented in a way that is understandable and digestible, with adequate time to ask questions, consider the uncertainties, and revisit the decision as new information becomes available.

We also have a responsibility as a field to be as transparent as possible about both successes and failures. Companies and researchers should share safety information openly and promptly so that we can learn collectively and avoid repeating preventable mistakes. No individual researcher or company benefits from the field learning the same safety lesson multiple times because information was not shared.

Ultimately, every powerful therapy carries risks. The more important question is whether we understand those risks well enough to determine when they are justified by the potential benefit for a particular child and whether we are acting as responsible stewards of the trust that families place in us when they choose to participate in these therapies and trials.

There is no simple answer to [the question of] how dangerous is gene therapy. One answer is that yes, they are safe! Hundreds of thousands of patients around the world have received gene therapies without serious adverse events with clear clinical benefits. With that as the background, there have been serious adverse events in some patients, even resulting in deaths. The field tries to learn as much as they can from these events so as to design gene therapies in the future that avoid them.

Unfortunately, in certain circumstances, there can be a lack of transparency about what happened and that makes it difficult for the field to learn and improve and right now, that is the case with the two deaths reported from China. We really don’t know enough of the details to understand what happened scientifically. It is essential that all patients who are considering receiving a gene therapy, whether investigational or approved, receive unbiased, informed information about the potential risks and benefits.

For any new type of therapy, including gene therapies, there are both known and unknown risks but as a category I would not say that gene therapies are inherently more risky than other treatments. In the United States, there is in fact, a very rigorous process to assess safety of a gene therapy before it is ever given to a person. And many gene therapies are developed for diseases that inevitably lead to early death and so the risk/benefit analysis that gene therapy developers, regulators, clinicians, patients and their families undertake is shaped by that.

There is incomplete information about the two cases in China, but from what we do know, the case of death from the brain delivery of AAV with a gene editor would probably not have happened in the U.S. There were numerous problematic scientific, medical, ethical, and conflict of interest issues in this case, any of one of which should have stopped the therapy and probably would have in the U.S. with the redundant safeguards that established hospitals and research institutions have in place, including FDA oversight. The case of the death using AAV to deliver a gene editor for muscular dystrophy might have happened even in the U.S. If it had happened in the U.S. though, there probably would have been much more scientific transparency about the event. Again, details are lacking but it does seem like the company did the right thing by stopping the trial after the death.

The promise and premise of gene therapy is exceptional. The reality that a one-time cure can be provided for a disorder that is utterly devastating like spinal muscular atrophy has generated considerable excitement from a patient and industry standpoint. Therapies stemming from adeno-associated virus delivery of genes or CRISPR machinery can provide hope to individuals who often have no other treatment options. However, the risks can be sobering. Consequently, the field of gene therapy has proceeded cautiously. The regulatory oversight to obtain FDA approval for the first approved gene therapy for inherited retinal disease was necessarily challenging. This has led to a stepwise progression to other conditions that are debilitating and can lead to early mortality. As additional, perhaps less devastating conditions are considered for therapy, the risks should be carefully considered.

Vigilant monitoring of the many patients now enrolled in trials has arguably made gene therapy safer than ever. Toxicity appears to be driven by the amount of virus (the dose), the strength of expression (the promoter), and the location of delivery. Acute consequences after delivery have led to widespread administration of steroids to prevent immune responses, for example. Less concerning has been the nature of the gene itself or the CRISPR guide RNA sequence. Adopting the FDA’s Platform Technology Designation Program could help expand the number of patients eligible for treatment.

Transparency in trial design is essential and is critical in understanding clinical complications, many of which can appear months or even decades after a therapy is initiated. Obscuring details of adverse outcomes does a disservice to the field and the brave participants in gene therapy trials. Moreover, while acute decline is easy to identify, medium or long-term effects are still unknown and can include complications such as a brain tumor that occurred in an individual several years after viral delivery. Work from gene therapy delivered in mouse neonates has revealed consistent liver cancers late in life—something that is still to be determined in humans. Understanding how these cancers arise and can be mitigated is of paramount importance.

The accounts that the parents provide for the young girl that died in China are particularly heart wrenching. We often hear glowing testimonials of patients with clinical improvements reaching developmental milestones. Balancing these stories with treatment failures can provide context to patients to better inform their consent.

The investigator-initiated trial in China does provide flexibility in that there may be rare conditions or private citizens who are inspired to find cures for loved ones and have the financial resources to provide funding for studies. Parallel examples can take place in the States, especially for terminal diseases through expanded access or compassionate use authorization. However, extensive discussions and consultations need to take place with stakeholders because of the relatively nascent field. The race to be first for a strategy or condition needs to be weighed extensively with associated complications.

One thing is clear is that designing a bespoke therapy for a living individual can create a sense of urgency that can drive innovation and achieve therapeutic success, yet it underscores the need for independent oversight.

Many serious genetic diseases happen when both copies of a gene, one from each parent, are mutated, so there’s no working backup copy left to make the protein we need. These diseases are individually rare, so there’s not much commercial incentive to develop treatments for them, and gene therapy is often the only option. Sometimes it’s the only thing that lets a child survive to adulthood.

Gene therapy delivers a healthy copy of the gene into cells so they can make the missing protein themselves. The main bottleneck is delivery efficiency, and the best tool we have for that is Adeno-Associated Virus, or AAV. For treatment, we use a stripped-down version of the virus so it can no longer replicate, so it can’t spread or cause disease, but it’s still efficient at getting genetic cargo into cells and keeping it active for years. That combination of safety, efficiency, and durability has made AAV the backbone of the gene therapy field, with hundreds of clinical trials and thousands of patients treated over the past decade.

But as more trials happen, rare adverse events are accumulating and getting attention. There are two main ways this happens. First, the immune system can recognize AAV as foreign and, in rare cases, mount a strong enough response to damage healthy tissue, which is dangerous if that tissue is the heart, lungs, liver, or kidneys. Second, in less than 1% of cases, a cell mistakenly inserts the viral DNA into its own genome during routine repair. Most of the genome isn’t sensitive to this, but on rare occasions the insertion switches on a cancer-causing gene and leads to tumor formation. Both risks go up with higher viral doses, which is why figuring out the right dose, frequency, and tissue target is central to clinical trial design and why the first phase of any trial is entirely focused on safety before efficacy is ever tested.

That safety-first process is the right one, but it’s also slow, often taking years. For patients with fatal, fast-progressing diseases, that timeline itself becomes a risk, and some lives have been lost to the wait, not the therapy. This is a genuinely hard tradeoff for policymakers: more safety checks can mean missing the window to save someone, but fewer checks can mean real danger too. Rather than looser regulation, what might help is faster, well-resourced external review for these high-stakes cases.

Meanwhile, a lot of research has gone into alternatives. Small RNAs like siRNA and antisense oligonucleotides can’t integrate into DNA, so they avoid that risk, and both are already FDA approved for certain conditions, though they mainly reduce gene activity rather than replace a missing gene outright. mRNA and self-amplifying RNA can deliver a whole gene the way AAV does, but current versions are less stable or don’t last as long. None fully match AAV’s combination of safety, efficiency, and longevity yet, but given the pace of progress in RNA medicine, real alternatives are probably coming. For now, AAV is still the most powerful tool we have, and it’s only getting safer.

The vast majority of genetic diseases result from disruption of a single gene, yet their consequences can be devastating and lead to premature death. Despite the severity of these conditions, therapeutic options for most rare genetic diseases remain extremely limited or entirely absent. Gene therapy has changed this trajectory and is already delivering life-changing, and in some cases potentially curative, treatments for diseases that were previously considered untreatable.

Gene therapies are not inherently more dangerous than other treatments, but they present unique risks because they are designed to produce long-lasting, and often irreversible, biological changes. The risk also varies considerably depending on the therapeutic platform, the gene being targeted, the delivery system, the dose, and the underlying disease. Importantly, the risks of treatment must always be weighed against the risks of the disease itself. For a child with a rapidly progressive and ultimately fatal genetic disorder, delaying treatment can itself have irreversible consequences.

Alongside remarkable successes, the gene therapy field has faced significant setbacks. Adverse events understandably receive substantial attention because of the novelty and potential risks of these technologies. However, disproportionate attention to individual events can also create periods of hesitation that slow scientific and regulatory progress for patients who have few or no alternatives. Importantly, adverse events have provided critical lessons that have strengthened the field. Past complications have driven technological innovation, improved our understanding of gene delivery and immune responses, and led to increasingly rigorous regulatory oversight and clinical monitoring.

The recently reported deaths of children receiving gene therapies in China are therefore an important reminder that rapid development cannot come at the expense of appropriate safety evaluation and careful patient monitoring. Serious adverse events could occur anywhere, including in the United States, because no medical intervention is without risk. However, rigorous preclinical testing, regulatory review, carefully designed clinical trials, independent safety oversight, appropriate patient selection, and long-term follow-up are essential safeguards designed to minimize those risks and identify complications as early as possible.

At the same time, requiring every new gene therapy to repeat extensive testing when substantial safety knowledge already exists for a well-characterized platform can unnecessarily delay access. The appropriate path forward is a risk-based development paradigm that leverages prior knowledge while maintaining targeted, product-specific safety assessments. Our responsibility is not to eliminate all risk, which is impossible, but to ensure that risks are scientifically justified, carefully monitored, and proportionate to the severity and urgency of the disease. For children facing otherwise incurable and rapidly progressive disorders, both excessive risk and excessive delay can have profound consequences.

Gene therapy has made remarkable strides in improving the prognosis for children with rare diseases that, until recently, had few or no effective treatment options. We have made enormous progress in safety, oversight, and trial design since the tragic death of Jesse Gelsinger in 1999, but these are still novel therapeutics, and there are risks—both short- and long-term—that we may not yet fully understand.

For some families, it is entirely reasonable to accept a degree of uncertainty when the potential benefit is the opportunity to substantially alter the course of a severe childhood disease. For other families, the possibility of causing an unforeseen or irreversible downstream harm may feel unacceptable. I don’t think there is one right answer. Families have different values, different circumstances, and different tolerances for uncertainty, particularly when making decisions on behalf of a child.

That is why I believe ongoing informed consent and, when appropriate, assent need to be at the center of gene therapy trials. Consent for a potentially high-risk, first-in-human or otherwise experimental therapy cannot be a single event in which a family signs a document. It should be an ongoing conversation. Families need information presented in a way that is understandable and digestible, with adequate time to ask questions, consider the uncertainties, and revisit the decision as new information becomes available.

We also have a responsibility as a field to be as transparent as possible about both successes and failures. Companies and researchers should share safety information openly and promptly so that we can learn collectively and avoid repeating preventable mistakes. No individual researcher or company benefits from the field learning the same safety lesson multiple times because information was not shared.

Ultimately, every powerful therapy carries risks. The more important question is whether we understand those risks well enough to determine when they are justified by the potential benefit for a particular child and whether we are acting as responsible stewards of the trust that families place in us when they choose to participate in these therapies and trials.

There is no simple answer to [the question of] how dangerous is gene therapy. One answer is that yes, they are safe! Hundreds of thousands of patients around the world have received gene therapies without serious adverse events with clear clinical benefits. With that as the background, there have been serious adverse events in some patients, even resulting in deaths. The field tries to learn as much as they can from these events so as to design gene therapies in the future that avoid them.

Unfortunately, in certain circumstances, there can be a lack of transparency about what happened and that makes it difficult for the field to learn and improve and right now, that is the case with the two deaths reported from China. We really don’t know enough of the details to understand what happened scientifically. It is essential that all patients who are considering receiving a gene therapy, whether investigational or approved, receive unbiased, informed information about the potential risks and benefits.

For any new type of therapy, including gene therapies, there are both known and unknown risks but as a category I would not say that gene therapies are inherently more risky than other treatments. In the United States, there is in fact, a very rigorous process to assess safety of a gene therapy before it is ever given to a person. And many gene therapies are developed for diseases that inevitably lead to early death and so the risk/benefit analysis that gene therapy developers, regulators, clinicians, patients and their families undertake is shaped by that.

There is incomplete information about the two cases in China, but from what we do know, the case of death from the brain delivery of AAV with a gene editor would probably not have happened in the U.S. There were numerous problematic scientific, medical, ethical, and conflict of interest issues in this case, any of one of which should have stopped the therapy and probably would have in the U.S. with the redundant safeguards that established hospitals and research institutions have in place, including FDA oversight. The case of the death using AAV to deliver a gene editor for muscular dystrophy might have happened even in the U.S. If it had happened in the U.S. though, there probably would have been much more scientific transparency about the event. Again, details are lacking but it does seem like the company did the right thing by stopping the trial after the death.

The promise and premise of gene therapy is exceptional. The reality that a one-time cure can be provided for a disorder that is utterly devastating like spinal muscular atrophy has generated considerable excitement from a patient and industry standpoint. Therapies stemming from adeno-associated virus delivery of genes or CRISPR machinery can provide hope to individuals who often have no other treatment options. However, the risks can be sobering. Consequently, the field of gene therapy has proceeded cautiously. The regulatory oversight to obtain FDA approval for the first approved gene therapy for inherited retinal disease was necessarily challenging. This has led to a stepwise progression to other conditions that are debilitating and can lead to early mortality. As additional, perhaps less devastating conditions are considered for therapy, the risks should be carefully considered.

Vigilant monitoring of the many patients now enrolled in trials has arguably made gene therapy safer than ever. Toxicity appears to be driven by the amount of virus (the dose), the strength of expression (the promoter), and the location of delivery. Acute consequences after delivery have led to widespread administration of steroids to prevent immune responses, for example. Less concerning has been the nature of the gene itself or the CRISPR guide RNA sequence. Adopting the FDA’s Platform Technology Designation Program could help expand the number of patients eligible for treatment.

Transparency in trial design is essential and is critical in understanding clinical complications, many of which can appear months or even decades after a therapy is initiated. Obscuring details of adverse outcomes does a disservice to the field and the brave participants in gene therapy trials. Moreover, while acute decline is easy to identify, medium or long-term effects are still unknown and can include complications such as a brain tumor that occurred in an individual several years after viral delivery. Work from gene therapy delivered in mouse neonates has revealed consistent liver cancers late in life—something that is still to be determined in humans. Understanding how these cancers arise and can be mitigated is of paramount importance.

The accounts that the parents provide for the young girl that died in China are particularly heart wrenching. We often hear glowing testimonials of patients with clinical improvements reaching developmental milestones. Balancing these stories with treatment failures can provide context to patients to better inform their consent.

The investigator-initiated trial in China does provide flexibility in that there may be rare conditions or private citizens who are inspired to find cures for loved ones and have the financial resources to provide funding for studies. Parallel examples can take place in the States, especially for terminal diseases through expanded access or compassionate use authorization. However, extensive discussions and consultations need to take place with stakeholders because of the relatively nascent field. The race to be first for a strategy or condition needs to be weighed extensively with associated complications.

One thing is clear is that designing a bespoke therapy for a living individual can create a sense of urgency that can drive innovation and achieve therapeutic success, yet it underscores the need for independent oversight.

Many serious genetic diseases happen when both copies of a gene, one from each parent, are mutated, so there’s no working backup copy left to make the protein we need. These diseases are individually rare, so there’s not much commercial incentive to develop treatments for them, and gene therapy is often the only option. Sometimes it’s the only thing that lets a child survive to adulthood.

Gene therapy delivers a healthy copy of the gene into cells so they can make the missing protein themselves. The main bottleneck is delivery efficiency, and the best tool we have for that is Adeno-Associated Virus, or AAV. For treatment, we use a stripped-down version of the virus so it can no longer replicate, so it can’t spread or cause disease, but it’s still efficient at getting genetic cargo into cells and keeping it active for years. That combination of safety, efficiency, and durability has made AAV the backbone of the gene therapy field, with hundreds of clinical trials and thousands of patients treated over the past decade.

But as more trials happen, rare adverse events are accumulating and getting attention. There are two main ways this happens. First, the immune system can recognize AAV as foreign and, in rare cases, mount a strong enough response to damage healthy tissue, which is dangerous if that tissue is the heart, lungs, liver, or kidneys. Second, in less than 1% of cases, a cell mistakenly inserts the viral DNA into its own genome during routine repair. Most of the genome isn’t sensitive to this, but on rare occasions the insertion switches on a cancer-causing gene and leads to tumor formation. Both risks go up with higher viral doses, which is why figuring out the right dose, frequency, and tissue target is central to clinical trial design and why the first phase of any trial is entirely focused on safety before efficacy is ever tested.

That safety-first process is the right one, but it’s also slow, often taking years. For patients with fatal, fast-progressing diseases, that timeline itself becomes a risk, and some lives have been lost to the wait, not the therapy. This is a genuinely hard tradeoff for policymakers: more safety checks can mean missing the window to save someone, but fewer checks can mean real danger too. Rather than looser regulation, what might help is faster, well-resourced external review for these high-stakes cases.

Meanwhile, a lot of research has gone into alternatives. Small RNAs like siRNA and antisense oligonucleotides can’t integrate into DNA, so they avoid that risk, and both are already FDA approved for certain conditions, though they mainly reduce gene activity rather than replace a missing gene outright. mRNA and self-amplifying RNA can deliver a whole gene the way AAV does, but current versions are less stable or don’t last as long. None fully match AAV’s combination of safety, efficiency, and longevity yet, but given the pace of progress in RNA medicine, real alternatives are probably coming. For now, AAV is still the most powerful tool we have, and it’s only getting safer.

The vast majority of genetic diseases result from disruption of a single gene, yet their consequences can be devastating and lead to premature death. Despite the severity of these conditions, therapeutic options for most rare genetic diseases remain extremely limited or entirely absent. Gene therapy has changed this trajectory and is already delivering life-changing, and in some cases potentially curative, treatments for diseases that were previously considered untreatable.

Gene therapies are not inherently more dangerous than other treatments, but they present unique risks because they are designed to produce long-lasting, and often irreversible, biological changes. The risk also varies considerably depending on the therapeutic platform, the gene being targeted, the delivery system, the dose, and the underlying disease. Importantly, the risks of treatment must always be weighed against the risks of the disease itself. For a child with a rapidly progressive and ultimately fatal genetic disorder, delaying treatment can itself have irreversible consequences.

Alongside remarkable successes, the gene therapy field has faced significant setbacks. Adverse events understandably receive substantial attention because of the novelty and potential risks of these technologies. However, disproportionate attention to individual events can also create periods of hesitation that slow scientific and regulatory progress for patients who have few or no alternatives. Importantly, adverse events have provided critical lessons that have strengthened the field. Past complications have driven technological innovation, improved our understanding of gene delivery and immune responses, and led to increasingly rigorous regulatory oversight and clinical monitoring.

The recently reported deaths of children receiving gene therapies in China are therefore an important reminder that rapid development cannot come at the expense of appropriate safety evaluation and careful patient monitoring. Serious adverse events could occur anywhere, including in the United States, because no medical intervention is without risk. However, rigorous preclinical testing, regulatory review, carefully designed clinical trials, independent safety oversight, appropriate patient selection, and long-term follow-up are essential safeguards designed to minimize those risks and identify complications as early as possible.

At the same time, requiring every new gene therapy to repeat extensive testing when substantial safety knowledge already exists for a well-characterized platform can unnecessarily delay access. The appropriate path forward is a risk-based development paradigm that leverages prior knowledge while maintaining targeted, product-specific safety assessments. Our responsibility is not to eliminate all risk, which is impossible, but to ensure that risks are scientifically justified, carefully monitored, and proportionate to the severity and urgency of the disease. For children facing otherwise incurable and rapidly progressive disorders, both excessive risk and excessive delay can have profound consequences.

Gene therapy has made remarkable strides in improving the prognosis for children with rare diseases that, until recently, had few or no effective treatment options. We have made enormous progress in safety, oversight, and trial design since the tragic death of Jesse Gelsinger in 1999, but these are still novel therapeutics, and there are risks—both short- and long-term—that we may not yet fully understand.

For some families, it is entirely reasonable to accept a degree of uncertainty when the potential benefit is the opportunity to substantially alter the course of a severe childhood disease. For other families, the possibility of causing an unforeseen or irreversible downstream harm may feel unacceptable. I don’t think there is one right answer. Families have different values, different circumstances, and different tolerances for uncertainty, particularly when making decisions on behalf of a child.

That is why I believe ongoing informed consent and, when appropriate, assent need to be at the center of gene therapy trials. Consent for a potentially high-risk, first-in-human or otherwise experimental therapy cannot be a single event in which a family signs a document. It should be an ongoing conversation. Families need information presented in a way that is understandable and digestible, with adequate time to ask questions, consider the uncertainties, and revisit the decision as new information becomes available.

We also have a responsibility as a field to be as transparent as possible about both successes and failures. Companies and researchers should share safety information openly and promptly so that we can learn collectively and avoid repeating preventable mistakes. No individual researcher or company benefits from the field learning the same safety lesson multiple times because information was not shared.

Ultimately, every powerful therapy carries risks. The more important question is whether we understand those risks well enough to determine when they are justified by the potential benefit for a particular child and whether we are acting as responsible stewards of the trust that families place in us when they choose to participate in these therapies and trials.

There is no simple answer to [the question of] how dangerous is gene therapy. One answer is that yes, they are safe! Hundreds of thousands of patients around the world have received gene therapies without serious adverse events with clear clinical benefits. With that as the background, there have been serious adverse events in some patients, even resulting in deaths. The field tries to learn as much as they can from these events so as to design gene therapies in the future that avoid them.

Unfortunately, in certain circumstances, there can be a lack of transparency about what happened and that makes it difficult for the field to learn and improve and right now, that is the case with the two deaths reported from China. We really don’t know enough of the details to understand what happened scientifically. It is essential that all patients who are considering receiving a gene therapy, whether investigational or approved, receive unbiased, informed information about the potential risks and benefits.

For any new type of therapy, including gene therapies, there are both known and unknown risks but as a category I would not say that gene therapies are inherently more risky than other treatments. In the United States, there is in fact, a very rigorous process to assess safety of a gene therapy before it is ever given to a person. And many gene therapies are developed for diseases that inevitably lead to early death and so the risk/benefit analysis that gene therapy developers, regulators, clinicians, patients and their families undertake is shaped by that.

There is incomplete information about the two cases in China, but from what we do know, the case of death from the brain delivery of AAV with a gene editor would probably not have happened in the U.S. There were numerous problematic scientific, medical, ethical, and conflict of interest issues in this case, any of one of which should have stopped the therapy and probably would have in the U.S. with the redundant safeguards that established hospitals and research institutions have in place, including FDA oversight. The case of the death using AAV to deliver a gene editor for muscular dystrophy might have happened even in the U.S. If it had happened in the U.S. though, there probably would have been much more scientific transparency about the event. Again, details are lacking but it does seem like the company did the right thing by stopping the trial after the death.

The promise and premise of gene therapy is exceptional. The reality that a one-time cure can be provided for a disorder that is utterly devastating like spinal muscular atrophy has generated considerable excitement from a patient and industry standpoint. Therapies stemming from adeno-associated virus delivery of genes or CRISPR machinery can provide hope to individuals who often have no other treatment options. However, the risks can be sobering. Consequently, the field of gene therapy has proceeded cautiously. The regulatory oversight to obtain FDA approval for the first approved gene therapy for inherited retinal disease was necessarily challenging. This has led to a stepwise progression to other conditions that are debilitating and can lead to early mortality. As additional, perhaps less devastating conditions are considered for therapy, the risks should be carefully considered.

Vigilant monitoring of the many patients now enrolled in trials has arguably made gene therapy safer than ever. Toxicity appears to be driven by the amount of virus (the dose), the strength of expression (the promoter), and the location of delivery. Acute consequences after delivery have led to widespread administration of steroids to prevent immune responses, for example. Less concerning has been the nature of the gene itself or the CRISPR guide RNA sequence. Adopting the FDA’s Platform Technology Designation Program could help expand the number of patients eligible for treatment.

Transparency in trial design is essential and is critical in understanding clinical complications, many of which can appear months or even decades after a therapy is initiated. Obscuring details of adverse outcomes does a disservice to the field and the brave participants in gene therapy trials. Moreover, while acute decline is easy to identify, medium or long-term effects are still unknown and can include complications such as a brain tumor that occurred in an individual several years after viral delivery. Work from gene therapy delivered in mouse neonates has revealed consistent liver cancers late in life—something that is still to be determined in humans. Understanding how these cancers arise and can be mitigated is of paramount importance.

The accounts that the parents provide for the young girl that died in China are particularly heart wrenching. We often hear glowing testimonials of patients with clinical improvements reaching developmental milestones. Balancing these stories with treatment failures can provide context to patients to better inform their consent.

The investigator-initiated trial in China does provide flexibility in that there may be rare conditions or private citizens who are inspired to find cures for loved ones and have the financial resources to provide funding for studies. Parallel examples can take place in the States, especially for terminal diseases through expanded access or compassionate use authorization. However, extensive discussions and consultations need to take place with stakeholders because of the relatively nascent field. The race to be first for a strategy or condition needs to be weighed extensively with associated complications.

One thing is clear is that designing a bespoke therapy for a living individual can create a sense of urgency that can drive innovation and achieve therapeutic success, yet it underscores the need for independent oversight.

Many serious genetic diseases happen when both copies of a gene, one from each parent, are mutated, so there’s no working backup copy left to make the protein we need. These diseases are individually rare, so there’s not much commercial incentive to develop treatments for them, and gene therapy is often the only option. Sometimes it’s the only thing that lets a child survive to adulthood.

Gene therapy delivers a healthy copy of the gene into cells so they can make the missing protein themselves. The main bottleneck is delivery efficiency, and the best tool we have for that is Adeno-Associated Virus, or AAV. For treatment, we use a stripped-down version of the virus so it can no longer replicate, so it can’t spread or cause disease, but it’s still efficient at getting genetic cargo into cells and keeping it active for years. That combination of safety, efficiency, and durability has made AAV the backbone of the gene therapy field, with hundreds of clinical trials and thousands of patients treated over the past decade.

But as more trials happen, rare adverse events are accumulating and getting attention. There are two main ways this happens. First, the immune system can recognize AAV as foreign and, in rare cases, mount a strong enough response to damage healthy tissue, which is dangerous if that tissue is the heart, lungs, liver, or kidneys. Second, in less than 1% of cases, a cell mistakenly inserts the viral DNA into its own genome during routine repair. Most of the genome isn’t sensitive to this, but on rare occasions the insertion switches on a cancer-causing gene and leads to tumor formation. Both risks go up with higher viral doses, which is why figuring out the right dose, frequency, and tissue target is central to clinical trial design and why the first phase of any trial is entirely focused on safety before efficacy is ever tested.

That safety-first process is the right one, but it’s also slow, often taking years. For patients with fatal, fast-progressing diseases, that timeline itself becomes a risk, and some lives have been lost to the wait, not the therapy. This is a genuinely hard tradeoff for policymakers: more safety checks can mean missing the window to save someone, but fewer checks can mean real danger too. Rather than looser regulation, what might help is faster, well-resourced external review for these high-stakes cases.

Meanwhile, a lot of research has gone into alternatives. Small RNAs like siRNA and antisense oligonucleotides can’t integrate into DNA, so they avoid that risk, and both are already FDA approved for certain conditions, though they mainly reduce gene activity rather than replace a missing gene outright. mRNA and self-amplifying RNA can deliver a whole gene the way AAV does, but current versions are less stable or don’t last as long. None fully match AAV’s combination of safety, efficiency, and longevity yet, but given the pace of progress in RNA medicine, real alternatives are probably coming. For now, AAV is still the most powerful tool we have, and it’s only getting safer.

The vast majority of genetic diseases result from disruption of a single gene, yet their consequences can be devastating and lead to premature death. Despite the severity of these conditions, therapeutic options for most rare genetic diseases remain extremely limited or entirely absent. Gene therapy has changed this trajectory and is already delivering life-changing, and in some cases potentially curative, treatments for diseases that were previously considered untreatable.

Gene therapies are not inherently more dangerous than other treatments, but they present unique risks because they are designed to produce long-lasting, and often irreversible, biological changes. The risk also varies considerably depending on the therapeutic platform, the gene being targeted, the delivery system, the dose, and the underlying disease. Importantly, the risks of treatment must always be weighed against the risks of the disease itself. For a child with a rapidly progressive and ultimately fatal genetic disorder, delaying treatment can itself have irreversible consequences.

Alongside remarkable successes, the gene therapy field has faced significant setbacks. Adverse events understandably receive substantial attention because of the novelty and potential risks of these technologies. However, disproportionate attention to individual events can also create periods of hesitation that slow scientific and regulatory progress for patients who have few or no alternatives. Importantly, adverse events have provided critical lessons that have strengthened the field. Past complications have driven technological innovation, improved our understanding of gene delivery and immune responses, and led to increasingly rigorous regulatory oversight and clinical monitoring.

The recently reported deaths of children receiving gene therapies in China are therefore an important reminder that rapid development cannot come at the expense of appropriate safety evaluation and careful patient monitoring. Serious adverse events could occur anywhere, including in the United States, because no medical intervention is without risk. However, rigorous preclinical testing, regulatory review, carefully designed clinical trials, independent safety oversight, appropriate patient selection, and long-term follow-up are essential safeguards designed to minimize those risks and identify complications as early as possible.

At the same time, requiring every new gene therapy to repeat extensive testing when substantial safety knowledge already exists for a well-characterized platform can unnecessarily delay access. The appropriate path forward is a risk-based development paradigm that leverages prior knowledge while maintaining targeted, product-specific safety assessments. Our responsibility is not to eliminate all risk, which is impossible, but to ensure that risks are scientifically justified, carefully monitored, and proportionate to the severity and urgency of the disease. For children facing otherwise incurable and rapidly progressive disorders, both excessive risk and excessive delay can have profound consequences.

Gene therapy has made remarkable strides in improving the prognosis for children with rare diseases that, until recently, had few or no effective treatment options. We have made enormous progress in safety, oversight, and trial design since the tragic death of Jesse Gelsinger in 1999, but these are still novel therapeutics, and there are risks—both short- and long-term—that we may not yet fully understand.

For some families, it is entirely reasonable to accept a degree of uncertainty when the potential benefit is the opportunity to substantially alter the course of a severe childhood disease. For other families, the possibility of causing an unforeseen or irreversible downstream harm may feel unacceptable. I don’t think there is one right answer. Families have different values, different circumstances, and different tolerances for uncertainty, particularly when making decisions on behalf of a child.

That is why I believe ongoing informed consent and, when appropriate, assent need to be at the center of gene therapy trials. Consent for a potentially high-risk, first-in-human or otherwise experimental therapy cannot be a single event in which a family signs a document. It should be an ongoing conversation. Families need information presented in a way that is understandable and digestible, with adequate time to ask questions, consider the uncertainties, and revisit the decision as new information becomes available.

We also have a responsibility as a field to be as transparent as possible about both successes and failures. Companies and researchers should share safety information openly and promptly so that we can learn collectively and avoid repeating preventable mistakes. No individual researcher or company benefits from the field learning the same safety lesson multiple times because information was not shared.

Ultimately, every powerful therapy carries risks. The more important question is whether we understand those risks well enough to determine when they are justified by the potential benefit for a particular child and whether we are acting as responsible stewards of the trust that families place in us when they choose to participate in these therapies and trials.

There is no simple answer to [the question of] how dangerous is gene therapy. One answer is that yes, they are safe! Hundreds of thousands of patients around the world have received gene therapies without serious adverse events with clear clinical benefits. With that as the background, there have been serious adverse events in some patients, even resulting in deaths. The field tries to learn as much as they can from these events so as to design gene therapies in the future that avoid them.

Unfortunately, in certain circumstances, there can be a lack of transparency about what happened and that makes it difficult for the field to learn and improve and right now, that is the case with the two deaths reported from China. We really don’t know enough of the details to understand what happened scientifically. It is essential that all patients who are considering receiving a gene therapy, whether investigational or approved, receive unbiased, informed information about the potential risks and benefits.

For any new type of therapy, including gene therapies, there are both known and unknown risks but as a category I would not say that gene therapies are inherently more risky than other treatments. In the United States, there is in fact, a very rigorous process to assess safety of a gene therapy before it is ever given to a person. And many gene therapies are developed for diseases that inevitably lead to early death and so the risk/benefit analysis that gene therapy developers, regulators, clinicians, patients and their families undertake is shaped by that.

There is incomplete information about the two cases in China, but from what we do know, the case of death from the brain delivery of AAV with a gene editor would probably not have happened in the U.S. There were numerous problematic scientific, medical, ethical, and conflict of interest issues in this case, any of one of which should have stopped the therapy and probably would have in the U.S. with the redundant safeguards that established hospitals and research institutions have in place, including FDA oversight. The case of the death using AAV to deliver a gene editor for muscular dystrophy might have happened even in the U.S. If it had happened in the U.S. though, there probably would have been much more scientific transparency about the event. Again, details are lacking but it does seem like the company did the right thing by stopping the trial after the death.

The promise and premise of gene therapy is exceptional. The reality that a one-time cure can be provided for a disorder that is utterly devastating like spinal muscular atrophy has generated considerable excitement from a patient and industry standpoint. Therapies stemming from adeno-associated virus delivery of genes or CRISPR machinery can provide hope to individuals who often have no other treatment options. However, the risks can be sobering. Consequently, the field of gene therapy has proceeded cautiously. The regulatory oversight to obtain FDA approval for the first approved gene therapy for inherited retinal disease was necessarily challenging. This has led to a stepwise progression to other conditions that are debilitating and can lead to early mortality. As additional, perhaps less devastating conditions are considered for therapy, the risks should be carefully considered.

Vigilant monitoring of the many patients now enrolled in trials has arguably made gene therapy safer than ever. Toxicity appears to be driven by the amount of virus (the dose), the strength of expression (the promoter), and the location of delivery. Acute consequences after delivery have led to widespread administration of steroids to prevent immune responses, for example. Less concerning has been the nature of the gene itself or the CRISPR guide RNA sequence. Adopting the FDA’s Platform Technology Designation Program could help expand the number of patients eligible for treatment.

Transparency in trial design is essential and is critical in understanding clinical complications, many of which can appear months or even decades after a therapy is initiated. Obscuring details of adverse outcomes does a disservice to the field and the brave participants in gene therapy trials. Moreover, while acute decline is easy to identify, medium or long-term effects are still unknown and can include complications such as a brain tumor that occurred in an individual several years after viral delivery. Work from gene therapy delivered in mouse neonates has revealed consistent liver cancers late in life—something that is still to be determined in humans. Understanding how these cancers arise and can be mitigated is of paramount importance.

The accounts that the parents provide for the young girl that died in China are particularly heart wrenching. We often hear glowing testimonials of patients with clinical improvements reaching developmental milestones. Balancing these stories with treatment failures can provide context to patients to better inform their consent.

The investigator-initiated trial in China does provide flexibility in that there may be rare conditions or private citizens who are inspired to find cures for loved ones and have the financial resources to provide funding for studies. Parallel examples can take place in the States, especially for terminal diseases through expanded access or compassionate use authorization. However, extensive discussions and consultations need to take place with stakeholders because of the relatively nascent field. The race to be first for a strategy or condition needs to be weighed extensively with associated complications.

One thing is clear is that designing a bespoke therapy for a living individual can create a sense of urgency that can drive innovation and achieve therapeutic success, yet it underscores the need for independent oversight.