Harvard Medical School Assistant Professor Tang Xin on the Current Landscape and Future Prospects of Pediatric Brain Disorder Therapies

Editor’s Note: This article was written by Mia Wang Jinghan and republished with permission from VCBeat.

Can you explain it to people outside the industry in one sentence?Could you briefly introduce your field?

Our research focuses on drug development for brain disorders. We employ stem cells, gene editing, and genetic engineering techniques to establish human brain disease models, which are used to elucidate disease mechanisms and serve as the foundation for targeted drug discovery.

Could you please elaborate on the treatment modalities for pediatric brain diseases?A Brief Science PopularizationThen, among the relevant treatment modalities,What role has gene therapy played?

Pediatric brain disorders encompass a wide variety of conditions, a significant proportion of which are neurodevelopmental disorders—diseases arising during brain development. These differ from neurodegenerative diseases such as Alzheimer’s disease and Parkinson’s disease. Neurodegenerative diseases may result from various environmental factors accumulated over a lifetime, including diet, sleep, and stress. In contrast, neurodevelopmental disorders are primarily genetic in origin, typically manifesting within the first few years of life. Environmental factors generally play a minor role; these disorders are mainly caused by single-gene mutations or by the interplay of multiple genes.

Pediatric neurological disorders encompass a broad spectrum of conditions, including autism spectrum disorder, attention-deficit/hyperactivity disorder (ADHD), schizophrenia, and depression. Compared with many other diseases, pharmacological options and therapeutic regimens for neurological disorders are relatively limited, primarily for two reasons: first, drugs face significant challenges in crossing the blood-brain barrier; second, the brain is an organ of extreme complexity and precision, unlike the liver or kidneys. Consequently, management of neurological disorders often relies on medications to alleviate symptoms, while more severe conditions, such as epilepsy and brain tumors, necessitate highly precise neurosurgical interventions. As the pediatric brain is still developing, it remains a particularly vulnerable organ. Overall, current therapeutic capabilities are substantially insufficient, leaving a significant unmet medical need.

According to the latest statistics in the United States, approximately one in six children is affected by a neurodevelopmental disorder, indicating a relatively high prevalence and a broadly impacted population. These disorders manifest in various forms and with varying degrees of severity. More severe cases involve structural abnormalities such as cerebral palsy, brain malformations, and developmental deficits. Milder forms are characterized by impairments at the level of neural connectivity or gene expression.

Therapeutic options for neurodevelopmental disorders remain extremely limited to date. Although gene therapy holds the potential for a one-time curative treatment, significant technical barriers still hinder its application in treating neurological diseases. A major challenge lies in delivering gene therapies into the brain and ensuring adequate coverage of brain tissue. Given the diverse cell types within the brain, it is difficult to target therapy specifically to disease-relevant cell populations and, within those cells, to correct the various pathological issues caused by disease-causing genes. Each of these steps presents substantial difficulties. Despite the rapid advancements in gene therapy over the past decade, it has not yet reached a stage in clinical practice where it can accurately locate and modify target genes to treat diseases.

For the entire field in which you operate,What is the ultimate goal?You yourself in this fieldWhat contributions have been made?

Let us break down this objective: Neurodevelopmental disorders are, in most cases, issues arising from genetic mutations during a child’s development. These mutations may be present since the zygote stage; however, because children typically have limited interaction with the external environment during their first one to two years of life, symptoms of autism spectrum disorder are difficult to detect and are generally identified and diagnosed around the age of two or three.

If pharmacological or gene therapies are administered at age three or later, can they truly eradicate the underlying cause of the disease? More importantly, can they make up for the lost time? This is a critical question. Because the brain is continuously developing, whether recovery is possible after the window of brain development has closed remains a subject of debate. A traditional view holds that there is a brief critical period during which neurons develop, neural connections are formed, and various basic functions are established. Therefore, once brain development is disrupted by neurological disorders, any missed opportunities during this period may be irrecoverable.

Over the past decade, research on genetically engineered mouse models of disease has demonstrated that neurodevelopmental disorders are not irreversible. Correcting genetic mutations during adulthood can partially eliminate the underlying causes and alleviate symptoms. If we liken the entire brain to a guitar, the brain of a child with autism or other developmental disorders is not a shattered instrument, but rather one that is simply out of tune. Therefore, by making appropriate adjustments, it is possible to restore basic functions. A decade ago, this was merely a hypothesis; however, it has gradually become a consensus. Researchers have observed that many diseases can be reversed in mice and other preclinical models using various approaches, leading to the growing belief that this is a feasible endeavor.

In addition, early diagnosis is of paramount importance. In recent years, this has become increasingly feasible due to the declining cost of gene sequencing and the fact that many sequencing-based diagnostic tests are now covered by insurance. These tests enable effective diagnosis of neurodevelopmental disorders; once a genetic mutation is identified, its associated diseases can be determined, thereby significantly enhancing diagnostic accuracy.

Secondly, the extent to which this genetic mutation affects a child’s development, the types of interventions or treatments that may be needed at different ages, and their potential outcomes can now be predicted with considerable accuracy. We have compiled comprehensive big data on the subtypes of various diseases, their subsequent progression patterns, and the required therapies. Furthermore, through such genetic diagnostics, patients’ families can organize themselves into foundations or mutual aid groups, facilitating experience sharing, coordinated fundraising efforts, or support for relevant research institutions. In summary, over the past five to ten years, the trend toward earlier and more precise diagnosis and treatment has become increasingly pronounced.

Relevant Gene TherapyWill it raise any ethical concerns?

Gene therapy, in most cases, does not refer to treatment at the embryonic stage but rather to somatic cell therapy. The distinction between the two lies in the fact that embryonic therapy involves modifying the genes of sperm, eggs, or zygotes at the very beginning, thereby altering every cell in the body, including germ cells, so that all carry the genetic modifications. This is known as germline gene editing. In contrast, somatic cell gene editing targets specific diseases; for instance, in a patient with a brain disorder where the mutation may be present throughout the body, gene editing can be used to treat the disease by correcting only a small subset of mutated genes in a specific organ, leaving other cells in the body unchanged. As relevant technologies mature further, it may become possible to correct specific genes in the brain, which is a concept entirely distinct from germline editing.

Somatic cell editing affects only the individual patient and cannot be transmitted to future generations. In contrast, germline cell editing can lead to such outcomes, as the edits are present in all cells of the individual across all stages of development and are passed on to all descendants. Therefore, this is not merely a medical issue but arguably a public health concern, given that germline gene editing results in intergenerational inheritance. This raises highly complex social issues and may well exacerbate social inequities.

For instance, if a certain gene could cause a child to die before the age of three, we should certainly edit that gene if we have the capability. But if the gene merely results in short stature, should we still edit it? What if the goal is simply to ensure the child has a robust physique? Or what if parents desire red hair for their child, or wish to increase the likelihood of college admission? Such applications clearly cross ethical boundaries. Therefore, determining where to draw the line is a matter worthy of extensive discussion.

Artificial Intelligence, Which Has Gained Immense Popularity in Recent YearsIs it possible in certain aspectsSupporting the Treatment of Pediatric Brain Disorders?

Certainly. Artificial intelligence is a general-purpose technology with broad applicability, such as in the subtyping of various brain disorders. For instance, autism spectrum disorder encompasses a wide range of conditions, differentiated by severity and etiology. AI can play a significant role in distinguishing disease subtypes, including genotypic and phenotypic classifications. In these analytical domains, big data combined with population-level information can be leveraged to develop frameworks suitable for analysis.

These frameworks currently exist only within the experience and diagnostic reasoning of physicians. While doctors are highly experienced and skilled, their expertise is typically confined to specific subspecialties; they cannot possibly diagnose every type of disease. Indeed, even seasoned specialists may encounter rare or obscure conditions for the first time outside of textbook cases. In contrast, artificial intelligence has the potential to achieve a level of generalist competence. Given the vast number of disease subtypes, it is impractical for every hospital to have relevant specialists on staff. Therefore, using AI for initial diagnosis and triage, followed by referral to specialist clinics at appropriate hospitals, can provide tangible convenience to patients.

Pediatric Brain Diseases Compared to AdultsWhat are the unique characteristics?

Generally speaking, the treatment of pediatric diseases demands higher standards compared to adults. Due to children’s smaller body size, relatively weaker constitution, and ongoing brain development, drug administration must consider not only therapeutic efficacy but also potential side effects. Consequently, the safety standards for developing pediatric drug regimens are significantly more stringent. Furthermore, medications for neurological disorders differ from those for oncology. While tumor-suppressing drugs may achieve their goal within one or a few treatment courses, many neurological medications require long-term administration, which further elevates the requirements for drug safety. These medications must not produce severe side effects in pediatric patients with varying constitutions. Therefore, key challenges in pharmaceutical development include minimizing side effects, enabling effective blood-brain barrier penetration, and ensuring precise therapeutic delivery.