Primum Non Nocere: Should Gene Therapy Be Used to Prevent Potentially Fatal Disease but Enable Potentially Destructive Behavior?

Abstract

Aldehyde dehydrogenase 2 (ALDH2) deficiency constitutes one of the most common hereditary enzyme deficiencies, affecting 35% to 40% of East Asians and 8% of the world population. It causes the well-known Asian Alcohol Flush Syndrome, characterized by facial flushing, palpitation, tachycardia, nausea, and other unpleasant feelings when alcohol is consumed. It is also associated with a marked increase in the risk of a variety of serious disorders, including esophageal cancer and osteoporosis. Our recent studies with murine models have demonstrated that a one-time administration of an adeno-associated virus (AAV) gene transfer vector expressing the human ALDH2 coding sequence (AAVrh.10hALDH2) will correct the deficiency state and prevent alcohol-induced abnormalities of the esophagus and bone. If successful in humans, such strategy would reduce the increased risk-associated disorders such as esophageal cancer and osteoporosis, but also prevent the Asian Alcohol Flush Syndrome. This treatment thus raises ethical concerns: although it would potentially prevent fatal disease, it could also allow affected individuals to drink alcohol without suffering the Asian Alcohol Flush Syndrome and, hence, potentially enable personal destructive behavior. Here we explore the ethical arguments against the development of a gene therapy for ALDH2 deficiency and we find them wanting. We contend that development of such treatments is ethically appropriate and should be part and parcel of the solutions offered against the condition.

INTRODUCTION

A variety of human behaviors, including cigarette smoking and unprotected sex, are legal and socially acceptable, yet each is potentially associated with fatal disorders, with consequences to the individual, their families, and society. Cigarette smoking is linked to a high risk for lung cancer, chronic obstructive pulmonary disease, and cardiovascular disease.¹ Unprotected sex is associated with an increased risk for development of HIV-associated disease.

Despite the fact that both cigarette smoking and unprotected sex are personal destructive behaviors, the biomedical research community, pharmaceutical industry, and government have invested significant resources to develop therapies to treat the consequences of these behaviors, with generation of effective therapies for cigarette smoking-related disorders and drugs to protect against the risk of HIV-AIDS associated with unprotected sex.² But what if these therapies reduced not just the harms associated with smoking or unprotected sex but also made those activities perhaps more likely?

This is now an issue for the gene therapy community because we have developed a gene therapy that, at least in rodent models, prevents potentially fatal disease, including esophageal cancer and osteoporosis, in homozygotes with aldehyde dehydrogenase 2 (ALDH2) deficiency. However, it also prevents the so-called Asian Alcohol Flush Syndrome associated with it.

ALDH2, an autosomal gene coding for a mitochondria enzyme, is the key enzyme in ethanol metabolism.^3–6 ALDH2 is the second enzyme in the ethanol metabolism pathway, functioning to convert the toxic intermediate acetaldehyde to nontoxic acetate. The common genetic variant ALDH2*2 allele (E487K) results in a significant reduction in the capacity of the liver to metabolize the acetaldehyde, leading to accumulation of acetaldehyde in circulation even with moderate amounts of alcohol use.^7–14 Affecting 35% to 40% of East Asians and 8% of the world population, ALDH2 deficiency is one of the most common hereditary disorders.^13,15,16 ALDH2 deficiency causes the so-called Asian Alcohol Flush Syndrome. When ALDH2 homozygotes, and to a lesser extent heterozygotes, drink alcohol, the affected individual suffers highly unpleasant symptoms, including facial and body flushing, tachycardia, nausea, and headaches.^13,15–17

In addition to the Asian Alcohol Flush Syndrome, the ALDH2*2 variant allele is associated with a variety of neurologic, endocrine, cardiovascular, and dermatologic disorders, aberrant drug metabolism, with a marked increase in the risk of esophageal, head, and neck cancers, as well as osteoporosis.^13,15,16

Evidence from our recent animal studies has raised the possibility that a one-time administration of an adeno-associated virus (AAV) gene transfer vector expressing the human ALDH2 coding sequence (AAVrh.10hALDH2) can correct the deficiency state.^15,16 If successful in humans, such strategy would reduce the risks of serious and potentially fatal disorders associated with the ALDH2 deficiency state, but also eliminate for the individual the unpleasant alcohol-induced Asian Alcohol Flush Syndrome.

This presents an ethical dilemma: if we treat ALDH2 deficiency with gene therapy, we could reduce the increased risks of esophageal cancer and osteoporosis, but we also are enabling the treated ALDH2-deficient individual to drink alcohol without developing the Asian Alcohol Flush Syndrome, that is, we are potentially preventing fatal disease, but in doing so, we enable personal destructive behavior that has consequences not only for the individual, but also for society. This begs the question of whether the development and implementation of this gene-transfer strategy are ethically appropriate. In this article we explore this argument and find it wanting. We contend that nonmaleficence and justice considerations support the development of treatments to reduce the risk for ALDH2 deficiency-related health problems.

Gene therapy for ALDH2 deficiency

We have carried out two studies to assess the potential of AAV-mediated gene therapy to correct ALDH2 deficiency.^15,16 In both studies, we used AAVrh.10hALDH2, the AAVrh.10 serotype coding for human ALDH2 driven by the CAG constitutive promoter.

In the first study,¹⁸ we hypothesized that a one-time administration of an AAV gene transfer vector expressing the human ALDH2 coding sequence (AAVrh.10hALDH2) would correct the acute effects of ethanol consumption in the context of the deficiency state. AAVrh.10hALDH2 was administered intravenously to Aldh2 knockout (Aldh2^−/−) and Aldh2 E487K knockin homozygous (Aldh2^E487K+/+) mice. After acute ethanol ingestion, untreated ALDH2-deficient mice had elevated acetaldehyde levels and performed poorly in behavioral tests. In contrast, treated Aldh2^−/− and Aldh2^E487K+/+ mice had lower serum acetaldehyde levels and improved behavior. Thus, in vivo AAV-mediated ALDH2 therapy may reverse the acute effects of ethanol consumption on the deficiency state in ALDH2*2 individuals, eliminating the Asian Alcohol Flush Syndrome and reducing other acute effects of ethanol drinking.

In the second study,¹⁹ we hypothesized that AAVrh.10hALDH2 would also prevent serious chronic disorders associated with chronic ethanol ingestion in the ALDH2 deficiency state. AAVrh.10hALDH2 was administered intravenously to the same two models of ALDH2 deficiency, Aldh2 knockout homozygous (Aldh2^−/−) and knockin homozygous (Aldh2^E487K+/+) mice. Four weeks after vector administration, mice were given drinking water with 10% to 15% ethanol for 12 weeks. Compared with nonethanol drinking littermates, AAVrh.10hALDH2 administration prevented chronic ethanol-induced serum acetaldehyde accumulation and elevated liver malondialdehyde levels, loss of body weight, reduced hemoglobin levels, and reduced performance in locomotor activity tests. Importantly, there was marked reduction in the accumulation of esophageal DNA damage and DNA adducts, and development of osteopenia supporting the concept that AAVrh.10hALDH2 should be considered as a preventative therapy for the increased risk of chronic disorders associated with ALDH2 deficiency and chronic alcohol consumption.

Preventing ALDH2 deficiency-associated health problems: ethical issues

In general, the development of new strategies to reduce alcohol health-related harms should be welcome. Nonetheless, there might be some misgivings regarding the use of gene-transfer interventions to address the health harms associated with ALDH2 deficiency, that is, such treatments may be inconsistent with the principle of nonmaleficence and with justice requirements.

Nonmaleficence

Reservations regarding the development of treatments against ALDH2 deficiency might be grounded on the principle of nonmaleficence. From this perspective, treatment with gene therapy will ultimately be harmful to patients with ALDH2 deficiency because it could lead more affected individuals to begin drinking or to drink more than they would have otherwise. Because correcting ALDH2 deficiency would prevent the immediate pathophysiological effects of alcohol consumption, this would diminish people's incentives not to drink. Indeed, some evidence suggests that carrying the ALDH2*2 allele protects people against alcohol abuse and dependence.^20,21 Various studies have shown that possession of an ALDH2*2 allele is associated with lower likelihood of drinking regularly, lower quantity and frequency of alcohol use, lower frequency of heavy episodic drinking, and lower maximum number of drinks consumed in a 24-h period, lower levels of self-reported alcohol-related problems, and alcohol dependence.^20–25 Individuals with ALDH2*2 may learn to regulate their drinking as a response to their unpleasant physical responses to alcohol.²²

Further evidence supporting harm-increasing concerns regarding gene therapy of ALDH2 deficiency comes from the development of some approaches to address alcohol abuse. Such strategies include drugs that inhibit ALDH2 activities, such as disulfiram and cyanamide.²⁶ These drugs act as alcohol-drinking deterrents by causing the unpleasant experiences (e.g., nausea, tachycardia, and flushing) produced by the accumulation of acetaldehyde in the blood of treated patients and mimicking the Asian Alcohol Flush Syndrome suffered by ALDH2*2 allele carriers.²⁶ Similarly, some animal studies have explored the feasibility of inhibiting ALDH2 in the liver as the basis for developing a liver-targeted approach to reduce alcohol-seeking behavior.²⁷

Excessive alcohol use has well-known harmful effects on individuals' health, as well as significant negative social and economic consequences.²⁸ Alcohol consumption is the third major risk factor for disease burden. Worldwide, harmful alcohol use kills 3.3 million people and alcohol-related deaths account for 5.3% of all deaths. In fact, alcohol consumption has more impact on mortality than digestive system diseases (4.5%), diabetes (2.8%), traffic accidents (2.5%), tuberculosis (2.3%), and HIV/AIDS (1.8%), and hypertension (1.6%).²⁸ Alcohol use is also associated with >200 health conditions and causes of premature death including various cancers, neuropsychiatric disorders, cardiovascular diseases, liver diseases, suicide, and traffic accidents.

But alcohol abuse has devastating effects not only on the health of individuals who consume it, but also on families and societies. Harmful alcohol use constitutes a significant factor in a variety of criminal activities, including homicide, child abuse or neglect, sexual assault, and intimate partner violence.²⁹ Because alcohol impairs decision making, it also increases other risky behaviors, such as unprotected sex or sex with multiple partners, which can result in unintended pregnancy or sexually transmitted diseases.³⁰ Its use is associated with various social problems such as lost productivity, broken families, and unemployment.²⁸ Excessive alcohol use costs U.S. society >$249 billion annually.³¹

Given the serious negative effects of alcohol, if using gene therapy to treat ALDH2 deficiency were to lead to an increase in consumption, it would seem unethical to do so. However, although we recognize this is an appropriate concern, several considerations call it into question. First, although the evidence regarding the protective effects of ALDH2*2 allele and the use of ALDH2 inhibitor therapies that mimic the negative effects of ALDH2 deficiency is important, it is hardly conclusive. In fact, epidemiological studies show that alcohol dependence among populations of ALDH2*2 carriers who suffer the immediate negative effects of alcohol use is actually increasing.^18,32 This indicates that experiencing the effects of the Asian Alcohol Flush Syndrome is insufficient to prevent drinking. Moreover, evidence from animal studies shows that acetaldehyde has properties similar to addictive drugs and plays a central role in the rewarding, motivational, and addictive properties of alcohol.³³ Administration of a lentiviral vector encoding ALDH2 into the brain of rats bred for alcohol preference reduced ethanol consumption and binge drinking.³⁴ Hence, administration of AAVrh.10hALDH2 gene therapy may actually contribute to reducing addictive behaviors. This evidence taken together weakens concerns about the potential harms that can result from treating ALDH2 deficiency.

Second, the belief that using gene transfer to treat ALDH2 deficiency can lead those so treated to drink at all or to drink more betrays a disregard for the complex network of social and cultural forces that influence drinking and other addictive behaviors. Alcohol consumption is an activity considerably influenced by cultural and social forces.^15,35–37 Norms, both formal and informal, strongly affect drinking practices. Social networks, business practices, family relationships, leisure time norms, entertainment rules, not to mention public policies, all have significant influence in drinking patterns. In Japan, for instance, where 41% to 52% of people possess at least one ALDH2*2 allele,³⁸ norms about group harmony in business likely shape drinking behavior among business men. In the past few decades, and in parallel with the proliferation of business society in Japan, the percentage of heavy drinking men who are ALDH2 heterozygotes has risen substantially.¹⁵ Among Asian American youth, problematic drinking is also increasing, despite the fact that high percentage carry mutated ALDH2 alleles. Factors such as cognitive expectancies, acculturation, enculturation, experience of discrimination, mental health problems, and gender socialization limit the potential protective effects of these mutations against alcohol use.^32,39

Indeed, the complex cultural and social norms that influence alcohol consumption behavior can also shade light on people's motivations for drinking. As mentioned earlier, excessive alcohol use produces serious health harms and can have devastating consequences on a person's life. If avoiding all of these other well-documented and well-known negative effects fails to motivate people to refrain from drinking, or from harmful consumption, it seems implausible that suffering the Asian Alcohol Flush Syndrome alone will be more motivating. But if these other factors are sufficient to incentivize people not to drink excessively, then administration of AAVrh.10hALDH2 should pose no additional harms in this regard to individuals treated. After all, even if the treatment were successful in preventing the immediate physiological effects of drinking, such strategy will not prevent all of the other social and financial problems related to harmful alcohol consumption. This again undermines concerns about the potential harm-causing effects of treating ALDH2 deficiency.

Third, the argument that developing treatments against ALDH2 deficiency violates the principle of nonmaleficence actually fails to appropriately consider the risks and potential benefits of such treatments. Even if it were the case that treating ALDH2 deficiency resulted in larger number of affected individuals drinking or drinking more, the principle of nonmaleficence also requires that we weigh the potential harms of increase in alcohol consumption against the possible harms of ALDH2-deficient individuals who will drink not having access to treatment. When taking into account both the possible harms of treating and not treating carriers of the ALDH2*2 allele, we believe that failing to prevent the health harms associated with ALDH2 deficiency is actually what conflicts with such ethical principle.

As mentioned, ALDH2 deficiency is associated not only with the Asian Alcohol Flush Syndrome but also, more importantly, with increased risk of various cancers, including esophageal, and head and neck.^{13,15,16,40–42} Esophageal cancer is one of the deadliest cancers worldwide, with 5-year survival rates of 15.6% in the United States, 12.3% in Europe, and 31.6% in Japan.¹⁵ ALDH2 deficiency is also associated with a high risk of osteoporosis and hip fracture, potentially fatal disorders.^40,43–48 There is extensive evidence that ALDH2 deficiency is associated with an increased risk for a variety of other pathologies, including several psychiatric and personality disorders, neurological diseases such as Parkinson's and Alzheimer's, higher sensitivity to pain, diabetes, cardiac ischemia, stroke, and impaired metabolism of some drugs.^13,15,16 Importantly, this risk increase exists even in the context of moderate drinking.

Clearly, when assessing both the potential harms of treating ALDH2 deficiency and those of failing to treat carriers of the ALDH2*2 allele, it is by no means obvious that the former alternative would minimize harms more than the latter. Indeed, arguably, given that evidence shows that suffering the Asian Alcohol Flush Syndrome does not prevent people from drinking alcohol, and that the serious health harms related to ALDH2 deficiency occur even in the presence of moderate alcohol intake, refusing to minimize the health harms associated with ALDH2 deficiency certainly fails to respect the principle of nonmaleficence.

Justice

A second reservation regarding treatment of ALDH2 deficiency is grounded on justice concerns. From this perspective, one could argue that given that alcohol drinking is a voluntary chosen activity, societies should not use scarce resources to aid ALDH2*2 allele carriers who knowingly engage in a high-risk behavior. Hence, the argument goes, it is actually unjust to use scarce health resources to develop treatments such as gene-transfer strategies that aim to prevent harms resulting from autonomously chosen activities and for which people are responsible. However, although some theories of justice take responsibility to be central to judgments about fairness,^49–51 several considerations challenge the claim that it would be unjust to develop treatments to prevent the health consequences of ALDH2 deficiency.

First, the notion of responsibility that underlies this argument seems to be excessively broad. Without denying that people are at least in part responsible for choices that affect their health, and accepting that their choices can be autonomous, evidence regarding the social determinants of health^52,53 clearly shows the relevance that social conditions, policies, and norms have on choices that affect people's health and well-being. As indicated earlier, alcohol drinking is a social activity regulated by various norms that might make difficult or costly to avoid it. Moderate alcohol drinking is encouraged in a variety of contexts, from entertainment, to business relationships, to family gatherings. Aggressive alcohol marketing and promotion strategies, many targeting youth, also contribute to the normalization of alcohol use. Moreover, despite the well-known health and economic cost that result from harmful alcohol consumption, most countries are failing to develop and implement policies to address alcohol use as a risk factor for health.⁵⁴ If states are giving insufficient attention and resources to addressing alcohol use as a risk factor for health, it seems problematic to refuse to develop treatments that could ameliorate alcohol use harms.

Second, this understanding of responsibility for health is inconsistent with current societal norms. It might well be the case that no amount of alcohol is safe,⁵⁵ but few, if any, activities that human beings value are completely safe. For instance, driving, engaging in sexual relationships, drinking coffee, having a diet rich in animal fats, or hiking are all voluntary chosen activities that involve risks. Nonetheless, most societies do not refuse to develop treatments that minimize the risks of these activities or treat the harms they can cause.

Granted, people can employ risk-lowering strategies when engaging in driving, sex, hiking, and the like. But millions of people who drink alcohol do so in moderation rather than excessively. Condemning people to suffer the serious health harms associated with ALDH2 deficiency, when such risks could be reduced, seems a disproportionate punishment for their behavior. After all, alcohol consumption is, like driving, engaging in sexual relationships, or hiking, a legal activity usually chosen in autonomous ways. Indeed, as mentioned, it is an activity encouraged in most societies by a variety of norms and practices. Insofar as moderate drinking involves risks similar to other accepted and valued pursuits, singling out those who are at increased risk of harm from even moderate alcohol consumption for harsher judgment and treatment than similarly situated others is unjust.

Third, even if one were to argue that drinking alcohol is a questionable choice, we do not fail to try to prevent harm in other situations that involve people's questionable choices. We do not, for instance, refuse to treat people who fail to use seatbelts when driving, or those who ride a motorcycle without a helmet, smoke cigarettes, or who practice sex without protection. It is unlikely that most societies would choose not to treat people in these cases, even if doing so had some effect on the likelihood that people would engage less on those behaviors.

In fact, most societies consider harm reduction strategies even in cases that involve illicit behavior as at least a prudential, it not an ethical, imperative. For instance, people who inject drugs are at increased risk of HIV and hepatitis C virus (HCV) infection. Nonetheless, multiple international organizations, such as WHO, UNAIDS, and the UN Office on Drugs and Crime, and an increasing number of countries, support and fund various interventions that aim to prevent and treat HIV and HCV infections among people who inject drugs. Such interventions include needle and syringe programs, opioid substitution therapy, and condom distribution programs to prevent viral transmission to sexual partners.⁵⁶ Similarly, drug overdose kills more people than homicides and motor vehicle accidents combined.⁵⁷ Concerns about the devastating impacts of the opioid epidemic have also led many societies to promote public health interventions such as improving training and increasing the availability of naloxone as a way to reduce the harms of opioid addition.^58,59 If societies agree to promote various strategies with the aim of preventing harm even in cases that involve illicit behavior, it is difficult to see what ethical grounds could be offered to refuse the promotion of treatment strategies to prevent health-related harms associated with ALDH2 deficiency.

CONCLUSION

Although it is correct that gene therapy for ALDH2 deficiency will enable ALDH2-deficient individuals to drink without the unpleasant symptoms of the Asian Alcohol Flush Syndrome, it also has the potential to protect against potentially fatal disease. We believe that there are no compelling ethical reasons to reject the use of gene therapy to reduce the harm of alcohol consumption in ALDH2-deficient individuals. Ultimately failing to reduce harms associated with ALDH2 deficiency misunderstands the principle of nonmaleficence and amounts to a moralistic condemnation of alcohol consumption, a behavior that is socially sanctioned and encouraged, that constitutes an affront to justice.

Although we have focused this analysis on gene therapy for ALDH2 deficiency, as gene therapy is applied to an expanding range of human disorders, it is likely that parallel issues will arise in other contexts. We hope that the ethical analysis we have provided for gene therapy for ALDH2 deficiency will serve as a useful guide for gene therapists confronted with these issues in the future.

AUTHOR DISCLOSURE

R.G.C. is consultant and holds equity in LEXEO Therapeutics; LEXEO has an option with Weill Cornell Medical College to license the ALDH2 gene therapy program. R.G.C. is also a coinventor on a patent application related to this topic (U.S. application number 16/321,023).

FUNDING INFORMATION

These studies were supported, in part, by NIH R41AA027739 and R41AA028465.