Pharmaceuticals and Personal Care Products in the Environment: Scientific and Regulatory Issues
Pharmaceuticals and Personal Care Products in the Environment: Overarching Issues and Overview
Christian G. Daughton
Chief, Environmental Chemistry Branch, ESD/NERL, Office of Research and Development, Environmental Protection Agency, Las Vegas, NV 89119, USA; e-mail: firstname.lastname@example.org; 702-798-2207; fax 702-798-2142.
Summary: While the point-source emissions of pollutants from manufacturing waste streams have long been monitored and subject to controls, the environmental impact of the public's (i.e., the individual's) activities regarding the use of chemicals is more difficult to assess. Of particular question is the widespread release to sewage and surface/ground waters of pharmaceuticals and personal care products after their ingestion, external application, or disposal. Certain pharmaceutically active compounds (e.g., caffeine, nicotine, and aspirin) have been known for over 20 years to enter the environment, especially in populated geographic locales, by a variety of routes -- primarily via treated and untreated sewage effluent. A larger picture, however, has emerged only more recently, where it is evident that numerous personal care products (such as fragrances and sunscreens) and drugs from a wide spectrum of therapeutic classes can occur in the environment and drinking water (albeit at very low concentrations), especially in natural waters receiving sewage. As of 1999, nearly all ecological monitoring studies for pharmaceuticals and personal care products (informally referred to as "PPCPs" ) had been performed in Europe.
The full extent, magnitude, and ramifications of their presence in the aquatic environment are largely unknown. Whether PPCPs in the environment pose a risk to humans or wildlife is not known. Aquatic exposures are noteworthy in that they can be continuous and can result solely from the external presence of a stressor. A major issue yet to be addressed by ecotoxicological science is the impact on aquatic biota of stressors eliciting effects sufficiently subtle as to go unnoticed in real time - but whose cumulative impacts eventually yield recognizable outcomes but having no obvious cause. How might the cumulative impacts of unrecognizable subtle effects compare with the overt, acute impacts of known toxicants? Another important issue is the potential impact of inhibitors/inducers of multi-drug transport (efflux) systems (as well as the better-known cytochrome P450 monooxygenase isoforms) in aquatic biota.
"Emerging" Risks and Science Planning: One of the primary goals of the U.S. EPA's Office of Research and Development is to identify and foster investigation of previously "hidden" or potential environmental issues/concerns before they become critical ecological or human health problems -- pollution prevention being preferable to remediation/restoration (to minimize public cost and to minimize human and ecological exposure). A major route to achieving this end is to highlight potential environmental issues (an objective of this book), thereby fostering further research, and to compile and integrate the resulting data so that the scientific community and the public can reach informed decisions - ensuring that science provides the foundation for any eventual discussion/decisions (if any) to regulate or not regulate. When stakeholders are included in open discussions, diverse viewpoints are assimilated during early development/planning stages rather than having these viewpoints in conflict at later implementation stages. By being proactive, the process of ruling-in/ruling-out (with the use of fast-failure analysis) allows society to minimize confrontation and inefficiency, and thereby promote cooperation and conservation of economic resources.
Drivers of Ecological Change: Ecological change is effected by human activities primarily via three routes -- habitat disruption/fragmentation, alteration of community structure (e.g., introduction of alien/nuisance species), and chemical pollution. The scopes of the first two are highly delineated compared with that of chemical pollution. Previously unidentified or underappreciated aspects of chemical pollution involve chemical classes not before recognized as pollutants -- the chemicals that compose any of the long-established regulatory "lists" of pollutants are not all-inclusive of those having toxicological significance.
During the last three decades, the impact of chemical pollution has focused almost exclusively on conventional "priority pollutants" (it is important to recognize that the current lists of non-atmospheric priority pollutants were primarily established in the 1970s largely out of expediency -- that is, they could be measured with off-the-shelf analytical technology; these priority pollutants were NOT necessarily selected solely on the basis of risk but rather because they could also be measured at sufficiently low detection limits), especially on those collectively referred to as "persistent, bioaccumulative, and toxic" (PBT) pollutants or persistent organic pollutants (POPs). The "dirty dozen" is a ubiquitous, notorious subset of these (e.g., see UNEP at: http://www.chem.unep.ch/pops).
The Larger Picture and Risk Assessment: This diverse "historical" group of persistent chemicals (comprising mainly agricultural and industrial chemicals and synthesis by-products, heavily represented by highly halogenated organics), however, may only be one piece of a larger puzzle. This bigger picture, if it does exist, has been largely unattainable with respect to risk assessments. Many other chemical classes (those that can be loosely referred to as "unregulated bioactive pollutants" or "unassessed" pollutants) must also be considered to gain a better look at the larger puzzle. Pharmaceuticals and personal care products (PPCPs) can be viewed simply as an example of one set of the universe of environmental pollutants that have received little attention with respect to potential impact on either ecological or human health. Of particular concern are effects on non-target organisms (e.g., aquatic biota) -- effects caused by unintended/inadvertent sources (e.g., human exposure to PPCPs via drinking water bearing low levels of PPCPs that have survived upgrade treatment systems), and effects brought about by non-target receptors (binding at receptors for which a drug was not designed; while there are currently hundreds of target receptors around which drugs are designed, these possibly represent only but a small fraction of those that might exist, most of which represent the untold numbers of non-target receptors). The added risk of individual chemicals cannot be considered outside the context of all chemicals to which organisms are simultaneously or sequentially exposed.
Limitations of Toxicity Assessment: The primary endpoints of interest to risk assessment (and toxicology) have traditionally been human morbidity and mortality (viz. acute toxicity and carcinogenesis). Comparatively little attention has been paid to the universe of other endpoints through which toxicants can express their action in non-mammalians (esp. aquatic organisms) -- some of which might be very subtle but nonetheless could lead to unanticipated, insidious outcomes. Endpoints such as neurobehavioral, immunological, and endocrine homeostasis alterations could lead to previously unrecognized outcomes and to those that are not recognized using current assessment criteria. Subtle endpoints could also be effected by extremely low toxicant concentrations (toxicology's domain has traditionally focused on high doses - e.g., see 2), and responses are not necessarily dictated by linear dose-response extrapolations. For example, effects mediated via hormone-like compounds do not necessarily follow the monotonic sigmoid dose-response curve; alternate, unexpected responses, following U- and inverted-U-shaped curves, are also possible (3).
Any effects imparted to non-target organisms could differ between and among each class of therapeutics being that their receptors are different -- for example, among antimicrobials, endocrine modulators, selective serotonin re-uptake inhibitors (SSRIs), and antineoplastics. This fact, coupled with a large spectrum of species (both aquatic and terrestrial) that could be exposed, means that a very large array of toxicity screening procedures could be needed -- the prospects for a single apical assay are accordingly low. Accounting for wild-type drug-metabolism/transport polymorphisms further complicates any screening approach. Exposure to multiple classes (and to multiple members within individual classes) must also be considered. Additive, synergistic, and cumulative time-course effects cannot be fathomed unless our understanding of the aggregate "exposure universe" is expanded.
It is particularly important to keep in mind that any concern for ecological or human health risks posed by newly considered chemical classes must have a scientific basis not just in environmental occurrence and exposure, but ultimately, in measurable effects having significant outcomes. We must also be diligent to weigh any known untoward effects due to environmental release of PPCPs against their established benefits as consumer products.
Certain pharmaceutically active compounds (caffeine [e.g., 4], nicotine, and aspirin, among others) have been known for over 20 years to enter the environment via sewage. Only more recently, however, has a larger picture emerged -- where it is evident that numerous drugs and personal care products from a wide spectrum of therapeutic and consumer-use classes (many having potent biochemical activity) can be inadvertently released to the environment (albeit yielding very low concentrations). The routes of release include both direct (disposal and wastage from external application) and indirect (excretion, washing, and swimming) -- primarily via treated and untreated sewage effluent -- and also by terrestrial runoff (for example from confined animal feeding operations - CAFOs - and excreta from medicated pets) and by wind-borne drift of agriculturally applied antimicrobials to crops. Municipal/domestic sewage, more so than hospital sewage, is the major source for most drugs classes (but not all) and quantities. Only limited subsets of the large spectrum of commercial PPCPs have been documented in the environment so far. Many therapeutic classes have not yet been surveyed for. The scope of the issue is ill-defined -- the numbers and types of PPCPs occurring in surface and ground waters across a large number of watersheds and municipalities are only partly known. Few generalizations can be made. It is reasonable to surmise, however, that the occurrence of PPCPs in waters is not a new phenomenon -- it has only become more widely evident in the last decade because continually improving chemical analysis methodologies have lowered the limits of detection for a wide array of xenobiotics in environmental matrices and have greatly expanded the ability to detect polar compounds.
The fact that PPCPs can be introduced on a continual basis to the aquatic environment via treated and untreated sewage essentially imparts a quality of "persistence" to compounds that otherwise may not possess any inherent environmental stability -- simply because their removal/transformation (by biodegradation, hydrolysis, photolysis, etc.) is continually countered by their replenishment, establishing a pseudo-steady-state in a manner analogous to a bacterial chemostat.
The full extent, magnitude, and ramifications of the presence of PPCPs in the aquatic environment are largely unknown. The toxicological significance of PPCPs in the environment with regard to either humans or terrestrial wildlife is poorly understood for most PPCPs. Lack of comprehensive exposure data (which for the aquatic environment, in contrast to the terrestrial environment, can sometimes be inferred simply from occurrence/concentration data) is a critical limitation to the advancement of risk assessment.
Although there are no monitoring requirements in the U.S. (and most others countries) for PPCPs in water, a preponderance of ecological monitoring studies in Europe has laid the foundation for this field. The problem has been most well defined in Europe because of the higher confluence of sewage outfalls in municipal areas, lower per-capita water usage, and smaller stream flows, all resulting in higher discharge concentrations and lower dilution by receiving waters. In North America (esp. Canada and parts of rural U.S.), however, there may be a high incidence of untreated sewage discharge, and therefore potential for greater variety of chemical types and higher concentrations.
While the misuse (which often includes overuse) and subsequent direct and indirect release (e.g., from CAFOs) of antimicrobials (synthetic and natural antibiotics) and natural/synthetic estrogenic steroids to the environment has generated nearly all the controversy to date regarding pharmaceuticals as potential pollutants (e.g., use in livestock ), a plethora of other drug classes, bioactive metabolites and transformation products, and personal care products have yet to be examined. The environmental fate of drugs in the excreta of medicated pets (and in carcasses from drug-euthanized animals, such as with barbiturates) is another area of question. Just as with humans, numerous drug classes are administered to pets - both those approved by the U.S. FDA [see: FDA's "Green Book" of Approved Animal Drugs: http://dil.vetmed.vt.edu/] and those administered off-label by veterinarians and pet owners: including antimicrobials/parasitics, antihistamines, non-steroidal anti-inflammatories, "behavior modifiers"/antidepressants (an example is the recently approved use of clomipramine, a tricyclic antidepressant, for treating pet anxiety), CNS agents, gastric agents, heart drugs, hormones, antineoplastics; "off-label" refers to drug use directed at conditions for which the FDA has not deemed effective and therefore for which no information appears in the manufacturer's package insert. Terrestrial run-off of excreta could serve as a highly dispersed route of drug/metabolite introduction to both the terrestrial and aquatic environments.
The concerns regarding antimicrobials (esp. promotion of pathogenic resistance, e.g., see 6), hormones (esp. reproductive/developmental effects), and musks (ubiquitous persistence) in the environment are better established than for other PPCPs. Antimicrobials (esp. low concentrations) impose selective pressure for resistance (unabated growth), or more commonly, tolerance (temporary growth stoppage, but continued viability) among potentially pathogenic microorganisms; significantly (as discussed later), promotion of antimicrobial resistance is at least partly caused by the proliferation or over-expression of cellular "multidrug resistance" systems (efflux pump-mediated drug resistance), which serve to minimize the intracellular concentrations of toxicants. Moreover, the acquired resistance or tolerance can be permanent -- it is frequently genetically conserved, persisting in the absence of continued selective pressure by the antimicrobial. Antimicrobials also have the potential to alter microbial species diversity, leading to altered successional consequences.
In comparison, the ecological concerns that we might have (if any) regarding other classes of PPCPs are ill-defined. There are, however, known mammalian effects for certain drug classes that do point to the need for further investigation. For example, various antiepileptic drugs (e.g., phenytoin, valproate, and carbamazepine, the latter of which is frequently identified in sewage effluents) are becoming more recognized as human neuroteratogens, triggering extensive apoptosis in the developing brain (during critical developmental "exposure windows"), and consequently leading to general neurodegeneration (e.g., see 7,8 and references cited therein; 9); such data prompt the question of whether these compounds affect non-target species.
To better serve the public, effort needs to be applied to performing the appropriate and sufficient science (which includes establishing occurrence, exposure, susceptibility/effects) so that sound decisions can be made regarding human and ecological health -- decisions based on "sound" science (10) -- rather than on "policy" or arbitrary assumptions that result in measures that are overly or under-protective of ecological or human health and that thereby waste economic resources or, conversely, jeopardize health; note that establishment of "occurrence" of a PPCP includes not just its structural identification, but also its concentration, frequency, and geographic extent and distribution for a given environmental matrix.
Clarification & "Disclaimer": As a result of various science planning activities (within and outside government), confusion often develops with regard to the relationship between PPCPs and "endocrine disrupting compounds". Only a small subset of PPCPs are known or suspected of being direct-acting endocrine disrupting compounds (EDCs) (primarily synthetic steroids and other synthetic hormones, acting as hormone or anti-hormone modulating mimics -- agonists or antagonists, respectively). While many xenobiotics can have a wide range of ultimate, indirect effects on the endocrine system, few have direct effects (i.e., serve as immediate endocrine agonists/antagonists at the hormone-receptor level). As an example, the inhibition or induction (such as by triazine herbicides) of P450 aromatase can effect changes in androgen/estrogen ratios (11); this effect is not at the receptor level. It is important to note that PPCPs and direct-acting EDCs are NOT synonymous, and the toxicological concerns are usually totally different.
a.k.a: environmental estrogens, endocrine-disruptors, endocrine-modulators, estrogenic mimics, ecoestrogens, environmental hormones, xenoestrogens, hormone-related toxicants, hormonally active agents (phytoestrogens being a subset)
Furthermore, the endocrine system (and its interconnected signaling pathways) is extraordinarily complex and cannot be easily distilled to a simple issue of "disruption" or "modulation". While "disruptors" can act directly at the hormone-receptor level, they can also act indirectly via a plethora of alternative routes (e.g., nervous system, immune system, specific cellular transporter systems), most of which are not always considered in the scope of many of the current definitions of EDCs. Endocrine disruption, in general, is narrowly viewed as a reproductive/developmental issue. An excellent overview of EDCs can be found at the "Environmental Estrogens and other Hormones" web site (Bioenvironmental Research at Tulane and Xavier Universities): http://www.tmc.tulane.edu/ECME/ eehome. Whether EDCs represent a meaningful way to classify toxicants with respect to environmental risk -- whether they play a significant toxicological role in environmental exposure (especially for humans) -- continues to be being actively debated (e.g., see 12).
Another important point is that the focus of this overview is on the use and disposal of PPCPs as originating from the activities and actions of individuals and hospitals -- not those from the PPCP manufacturing sector, whose waste streams are more confined and controllable. Any overview of a potential environmental pollution issue cannot encompass all of the aspects of environmental chemistry, toxicology, and other disciplines. With that perspective, this book does not pretend to be anything other than a catalyst and guide for promoting further exploration of the literature by the reader.
Current State of Knowledge Regarding PPCPs in the Environment
PPCPs as Environmental Pollutants? PPCPs are a diverse group of chemicals that have received comparatively little attention as potential environmental pollutants. PPCPs comprise all drugs (whether available by prescription or "over-the-counter", including the new genre of "biologics" -- proteinaceous drugs), diagnostic agents (e.g., X-ray contrast media), "nutraceuticals" (bioactive food supplements such as huperzine A), and other consumer chemicals, such as fragrances (e.g., musks), sun-screen agents (e.g., methylbenzylidene camphor), and skin anti-aging preparations (e.g., retinoids). While already used in vast quantities, the consumption and usage of PPCPs is only expected to increase, driven partly by new, previously unforeseen applications of existing drugs -- examples include "chemopreventatives" (to reduce the chances of disease or slow its onset; e.g., tamoxifen for breast cancer; aspirin for colon cancer) -- and the growing practices of off-label prescribing and promotion of "physician samples" (the latter's combined U.S. 1999 quantities amounted to over $7B in equivalent retail sales) (13).
It is difficult to gain good estimates of total nationwide drug usage in the U.S. (ethical prescription and over-the-counter [OTC], both human and veterinary) and much more difficult to obtain regional usages. According to PhRMa's "Industry Profile 2000" report (http://www.phrma.org/publications/industry/profile00/ PhRMA_Tables.pdf), sales in the U.S. of ethical pharmaceuticals (by U.S. and foreign companies) during a 12-month period in 1999/2000 were roughly one-half of worldwide (ca. $100B of $200B total), veterinary sales being roughly 1-2% of total, and OTC sales ca. 2% of total. In 1997, the U.S. per capita expenditures on drugs were the fourth highest in the world ($319), with Belgium, Japan, and France being the top three, ranging from $321 to $351, respectively. Pharmaceutical sales in 1997 for the U.S. were 1.4% of gross domestic product. These figures reflect very large usage rates.
Many PPCPs are extremely bioactive compounds and are unwittingly introduced to the environment as complex mixtures via a number of routes -- especially sewage effluent (both treated and raw). Drugs differ from agrochemicals in that they often have multiple functional groups (including ionizable groups and more frequent and extensive fluorination) and often lower effective doses (sub mg/kg), complicating fate/transport modeling and lending an extra dimension to the analytical techniques required for monitoring; drug structure also spans the spectrum from very simple low-molecular weight to large, complex molecules. In contrast to the conventional PBTs, most drugs are usually neither bioaccumulative nor volatile. Personal care products, however, such as the musk fragrances and sun screen agents, tend to be more lipophilic.
To further pursue the issues related to PBTs, a good place to start is with the integrated set of tools developed under the EPA's Office of Pollution Prevention and Toxic Substances' PBT program -- the "PBT Profiler", which estimates environmental persistence, bioconcentration potential, and aquatic toxicity from chemical structure; tools such as this could perhaps be used to target potential PPCPs for monitoring. Indeed, such a modeling approach was used by Blok et al. (14) in an extensive "screening" evaluation for PBTs, where it was found that numerous, previously unevaluated candidates ("non-assessed" chemicals) are found to be pharmaceuticals.
The Published Literature: The world's published literature on the topic of PPCPs in the environment is disparate (spanning numerous unrelated disciplines), disjointed, and uneven in its coverage. It is also more difficult to locate and to examine in a holistic, comprehensive manner than for "conventional" pollutants. Several compendia and reviews appearing since 1999 (1, 15-18) serve as good places to start and have also served for much of the perspective for this overview. Also worth noting are the pre-1999 review articles cited in Daughton and Ternes (1). Finally, two brief articles are available that summarize some of the salient points from the seminal ACS meeting that served as the foundation for this book (19,20).
Being that most environmental scientists are not as familiar with many of the aspects of PPCPs as they are with POPs, a good place to access numerous databases and other information in addition to the Merck Index (21) or the Physicians Desk Reference (22), can be found in various world wide web databases provided through links at "Coreynahman.com Pharmaceutical News and Information" (http://www.coreynahman.com/index.html); for example, comprehensive compilations of physician materials for individual drugs in current use can be found at "RxList - The Internet Drug Index" (http://www.rxlist.com), "Scholz Health Care", or "Gold Standard Multimedia Clinical Pharmacology 2000" .. Information on new and anticipated drugs can be found at "Lexi-Comp's Clinical Reference Library" (http://www.lexi.com/new_drugs.htm [post-publication note -- URL changed to: http://www.lexi.com/web/newdrugs.jsp]).
Inter-Connectedness of Humans and the Environment: Perhaps more so than with any other class of pollutants, the occurrence of PPCPs in the environment highlights the intimate, inseparable, and immediate connection between the actions, activities, and behaviors of individual citizens and the environment in which they live. PPCPs, in contrast to other types of pollutants, owe their immediate origins in the environment directly to their worldwide, universal, frequent, highly dispersed, and individually small but cumulative usage by multitudes of individuals -- as opposed to the larger, highly delineated industrial manufacturing/usage of most high-volume synthetic chemicals. PPCPs enjoy true worldwide usage and concomitant potential for discharge to the environment. Their introduction to the environment has no geographic boundaries or climatic-use limitations as do many other synthetic chemicals. They are discharged to the environment wherever people (having access to medication or medical care) live or visit, regardless of season.
Manufactured and used in large quantities worldwide (from several to hundreds of kilograms, to thousands of tons - for each individual substance), PPCPs comprise a diverse array of pollutants -- usage rates of many are on par with agrochemicals. Escalating introduction to the marketplace of new pharmaceuticals; expanding usage of existing drugs (antidepressants, stimulants, and other physchotropic drugs to modify child behavior; off-label prescribing for both humans and pets; conversion of established drugs to OTC status as a result of the growing "self-care" movement being only three examples); new uses for "retired" drugs (e.g., thalidomide for inflammation); and the growing practice of procuring prescription drugs from other countries, are each adding perhaps exponentially to the already large array and amounts of PPCP chemical classes, each with distinct and expanding modes of biochemical action (many of which are poorly understood, especially in wildlife, and some needing only very low concentrations to impart effects). It is also interesting to note that some chemicals serve double duty as both existing/experimental drugs and as pest-control agents (e.g., 4-aminopyridine: an avicide and experimental multiple sclerosis drug; warfarin: a rat poison and anticoagulant; triclosan: general biocide and gingivitis agent used in toothpaste; azacholesterols: antilipidemic drugs and avian/rodent reproductive inhibitors [e.g., Ornitrol]; certain antibiotics: used for orchard pathogens; acetaminophen: an analgesic and useful for control of brown tree snakes); the potential significance of these alternative uses as sources for environmental release has never been explored. Usage of certain drugs also continuously increases as a result of abusing prescription drugs for non-medical, "recreational" purposes and because of unneeded dispensing ("imprudent use" and unfounded patient demands or physician expectations). Another source of highly bioactive chemicals, but whose impact on the environment is totally unknown, is the clandestine manufacturing and use of illegal drugs -- which constitutes a potentially large but highly dispersed source of totally different pollutants through surreptitious manufacture and direct disposal; see http://www.streetdrugs.org/ and National Institute of Drug Abuse (NIDA) (available: http://www.nida.nih.gov/NIDAHome1.html) for comprehensive information regarding "recreational"/street drugs. Also note the concluding chapter in this book (23), which focuses specifically on illicit drugs.
Sewage and domestic wastes are the primary sources of PPCPs in the environment (posing concerns for drinking water supplies?): These bioactive compounds are continually introduced to the environment (primarily via surface and ground waters) from human and animal use largely through sewage treatment works systems (STWs), failed septic fields, leaking underground sewage conveyance systems, and wet-weather runoff -- either directly by bathing/washing/swimming (via discharge of externally applied PPCPs, such as fragrances or sun-screen agents, or those excreted in sweat) or indirectly by excretion in the feces or urine of unmetabolized parent compounds. Bioactive metabolites (including reconvertible conjugates) are also excreted. Excretion of parent drug and metabolites is partly a function of the age/health, timing of dose (diurnal aspects of absorption/transport/metabolism; gestational/developmental stage), and constitution of the gut (microbial contribution to endogenous metabolism can be significant). It should be noted, however, that pharmacokinetics is not the only predictor for the fraction of a drug that is excreted unmetabolized. Actual drug formulation has the potential to prevent full dissolution of a drug in the gut, leading to minimized uptake and therefore significantly increased excretion of the parent drug; for example, the use of certain excipients, such as stearic acid derivatives in tablets, is reported to sometimes greatly reduce the dissolution of a drug, resulting in excretion of undissolved pills (24). Disposal via municipal refuse serves as another route of introduction to the environment (e.g., via leaching to groundwater). Other routes to the environment include storm overflow events, residential "straight piping", and disposition of the massive quantities of drugs contributed to humanitarian assistance projects (e.g., 25).
While aspirin and caffeine have long been known to occur in sewage, only since the 1980s have other PPCPs been identified in surface and ground waters -- even certain drinking waters. As a recent example, low ng/L levels of clofibric acid, diazepam, and tylosin have been identified in Italian drinking water (26). Portions of the free excreted drugs and derivatives can escape degradation in municipal sewage treatment facilities. Removal efficiency in an STW or drinking water upgrade facility is a joint function of the drug's structure (including its stereochemistry, which can lead to chiral enrichment by preferential bioalterations) and the treatment technology employed (e.g., ozonation, chlorination, irradiation, carbon sorption). Treatment effectiveness would be expected to vary greatly across drug classes (as a function of chemical structure). For waste streams, one might expect especially poor removals for certain antimicrobials and antineoplastics (because of acute toxicity to various microbial species) and for highly sterically hindered compounds (e.g., iodinated contrast media); effectiveness could also be a function of the waste stream's origin (e.g., hospital waste versus domestic waste, especially with respect to higher concentrations and more acutely toxic drugs, such as genotoxicants used primarily at hospitals). Any metabolically conjugated derivatives released to the environment can eventually be converted back to the free parent drug; therefore, while not detectable as the parent compound, conjugates can serve as surreptitious reservoirs of additional toxicant, serving to regenerate the parent. General, overall STW effectiveness can fluctuate due to time of day (both composition and volume of sewage influent, the latter of which is largely a function of diurnal population activity, precipitation/runoff input, and industrial contributions) and season (treatment efficiency influenced by temperature and nutrient loads/physicochemical conditions, and dilution of effluent as a function of receiving water volume/flow). The intricacies involved with enhancing treatment effectiveness are complex, as exemplified by oxidative treatment of drugs in water by ozonation (27).
The low concentrations of individual PPCPs (possibly below the catabolic enzyme affinities of sewage microbiota), coupled with metabolic "novelty" to microorganisms (possibly an issue with newly introduced drugs), could lead to incomplete removal from STWs; removal efficiencies from STWs can span the entire spectrum -- from complete to ineffective. Introduction of many PPCPs to individual STWs is in the multi-g or -kg/day range, depending on the population served. While lipophilic PPCPs could partition to the solids (sludge), resulting in subsequent disposal in various ways on land (e.g., use of sewage sludge as agricultural fertilizer), little data is yet available; worth noting, however, is that living plants have the potential to bioaccumulate drugs (irrespective of polarity), giving the potential for subsequent transfer to animals and humans (28). While veterinary pharmaceuticals also are excreted (or washed from skin/hair), their route to the environment tends to be directly via wet-weather runoff (e.g., from CAFOs and medicated pets) from land to surface waters or ground waters, as well as directly via losses from aquaculture; the introduction of a wide array of antibiotics via aquaculture has been well documented, and these compounds show considerable persistence and a large spectrum of non-target effects (29), including phytotoxicty at ppb levels (30). Sometimes, the uptake of aquaculture medication is greater by nearby wild fin fish and shellfish than by the targeted cultured fish because the latter's diseased states minimizes uptake and absorption (31).
Whether PPCPs survive in natural waters sufficiently long down-gradient to be taken up in untreated drinking water, or whether they survive drinking water treatment, creating the potential for long-term exposure of humans, has received even less investigation than has environmental occurrence. Certain drugs/metabolites, however, have been documented in potable waters in Europe (see references cited in Daughton and Ternes ; 26). The extremely low concentrations (parts per trillion, ng/L), orders of magnitude below therapeutic threshold levels, might be expected to have minuscule (but still unknown) health consequences for humans, even for those who continually consume these waters over the course of decades; the primary concern, if any, would focus on those with heightened drug responses or the health-impaired (e.g., fetuses, infants, and children, or aged or diseased individuals). The estrogenicity of STW effluents in the UK (as measured by vitellogenin production in male fish - primarily a function of natural and synthetic steroid estrogens) has been shown to persist for several kilometers downstream of effluent discharges (32), maximizing the potential for drinking water uptake in rural areas that do not use upgrading facilities or that practice minimal treatment. It is important to keep in mind, however, that most drinking water is subjected to yet further treatment (upgrading), which lessens even further the chances of PPCP contamination. The issue of potable water contamination is more pertinent to situations where drinking water is not upgraded -- for example, when withdrawn directly from contaminated surface or ground waters (e.g., private wells); recharge of groundwater using treated sewage with residual PPCPs may pose a particular concern deserving further attention. The National Research Council (33) recently pointed to pharmaceuticals as one of several major chemical classes pollutants that had not received sufficient attention as potential water pollutants and recommended that they be considered among the larger universe of previously unevaluated pollutants for future versions of the Environmental Protection Agency's Drinking Water Contaminant Candidate List (CCL; see: http://www.epa.gov/OGWDW/ccl/cclfs.html).
In addition to the sources of drugs in the environment discussed here, one should reasonably expect to see a number of other routes of introduction of drugs to the environment with the advent of new technologies for discovery, manufacturing, and administration. For example, transgenic production of "functional food" drugs/vaccines (but mainly proteinaceous) by genetically altered plants (a.k.a. "molecular farming", "biopharming") could pose additional concerns for the unintended or uncontrollable release of drugs to the environment (e.g., see: 2nd International Molecular Farming Conference 1999 (29 August - 1 September 1999, London, Ontario, Canada).
Ubiquitous, persistent, and bioaccumulative? Very few PPCPs (primarily certain personal care products) and their metabolites have been found to display all the qualitites of conventional POPs. Many display one or two of these qualities, but few (such as musks) occur widely, persist, and bioaccumulate. Some (such as clofibric acid) occur at concentrations in surface waters on par with that of the widely recognized and ubiquitous organochlorine POPs (e.g., DDT, PCBs). Concentrations in natural surface waters (including oceans) generally range from ppb (µg/L) to ppt (ng/L). Some PPCPs are extremely persistent and introduced to the environment in very high quantities (e.g., see chapters in this book on musks and on polyiodinated X-ray contrast media). Others act as if they were indeed long-lived persistent pollutants simply because their source (sewage effluent) continually replenishes any removal caused by way of natural environmental processes (e.g., microbial degradation, photolysis, particulate sorption). Continual replenishment effectively sustains perpetual, multi-generational life-long exposures for aquatic organisms. Chiral enrichment (by selective biodegradation of chiral isomers) is another consideration. While most PPCPs display degradation in surface waters (preventing their accumulation but not full-time presence), their fate in ground waters (where the likelihood of persistence is maximized) is less understood (see chapter in this book by Drewes and Shore). As such, there is an imperative that we examine those possible environmental contamination situations that can be reversed only with difficulty. Groundwater recharge using treated wastewaters is one area that may need to be specifically scrutinized -- to ensure that contamination of these critical resources is minimized; extreme, unanticipated persistence in groundwater has been noted in some situations (e.g., pentobarbital; 34). Of particular noteworthiness is the almost complete absence of monitoring data on the presence (or absence) of PPCPs (other than musks) in fin- and shell-fish. While the polarity of most drugs precludes any concern about their bioconcentration in lipid, occurrence as adducts in various tissues cannot be excluded. Understanding of metabolic fates in aquatic biota is basically unknown.
A myriad of chemical classes, ranging from endocrine disruptors, antimicrobials, antidepressants ... to lipid regulators and synthetic musk fragrances: Excluding the antimicrobials and steroids (which include many members), over 50 individual PPCPs or metabolites (from more than 10 broad classes of therapeutic agents or personal care products) had been identified as of 1999 in environmental samples (mainly surface and ground waters) (see: Table I, summarized from Daughton and Ternes ); concentrations generally range from the low ppt- to ppb-levels. It is important to note that these only comprise a subset of those in wide use -- members of most classes have never been searched for in the environment (see: Table II).
Significantly, many of these compounds have no published aquatic toxicity data; some might have the potential for significant effects (e.g., antiepileptics, antineoplastics). Conversely, some PPCPs (such as the SSRI antidepressants, calcium-channel blockers, and efflux pump inhibitors) that do have published aquatic effects data (albeit limited) have yet to be surveyed in environmental samples (and therefore are not listed in Table I). Still others have great potential for profound aquatic effects but have neither the aquatic toxicological database nor any occurrence data (e.g., psychoactive agents and street drugs).
Majority of PPCP classes have no environmental survey data:
The number of PPCP classes (and individual members) that have been surveyed in various environmental samples (Table I) serve as a foundation that needs to be built upon. There are many other classes of drugs that have yet to be subjected to environmental surveys (see examples in Table II), and many of the most widely prescribed members of the classes in Table I have not been reported from environmental surveys. Note that while the literature is silent regarding these classes, it cannot be concluded whether this is because of an absence of data or because of a failure to report "data of absence"; it is important to note that investigators need to value reporting negative results. Many of these drugs are among the most widely prescribed in the U.S. (see: http://www.rxlist.com/top200.htm -- data compiled from American Druggist). Also note that the numerous ingredients in personal care products and the many bioactive compounds in "nutraceuticals" (e.g., the active ingredient[s] in St. John's wort) are not even considered here. Attention is only just beginning to be focused on the plethora of ingredients in personal care products. Phthalate esters is a case in point (35); mercury (from dental amalgam and toiletries) is another (36). The list of drugs newly identified in environmental samples continually expands -- new ones are continually added to the literature. For example, since Table II was compiled, several drugs on that list (plus others) have now been identified. For example, Zuccato et al. (26) identified in appreciable amounts up to hundreds of ng/L in Italian rivers of atenolol (antianginal, antihypertensive, beta blocker), ranitidine (histamine H2-receptor antagonist; e.g., Zantac), and furosemide (sulfonamide-type, loop diuretic). Snyder et al. (37, see chapter in this book) identified in lake water at the tens-to-hundreds of ng/L range a number or previously unreported drugs, including phenytoin, phenobarbital, primidone, hydrocodone, codeine, and others; some of these have also been reported by Möhle and Metzger (see chapter in this book). There is no reason to think that the incidence of discovery will not continue to grow; it is also important to note that the spectrum of PPCPs identified in one region or country can differ from those in another as a result of prescribing and usage patterns.
Environmental survey -- Data for a comprehensive approach are largely lacking: Environmental surveys for PPCPs (which analytes to target) can be guided somewhat by ranking the expected prevalence in STW effluents of each PPCP and using the ranked results to design a target-analyte approach. For personal care products, this can be accomplished by determining which of the myriad ingredients might be bioactive and then using industry production figures for combined members of each consumer-chemical "class". For medicinals, the objective is more complex. Production/consumption figures for individual locales are largely confidential or not available. Prescriptions filled and amounts consumed are difficult to acquire (for some general direction, refer to American Druggist, e.g., RxList Top 200: http://www.rxlist.com/top200.htm); usage figures for regional/local levels are nearly impossible to acquire (and the types of drugs can vary from region to region, country to country - for example, according to the age structure or usage conventions of the local population). Knowledge is needed of the excretion efficiency for the unaltered parent drug and its respective conjugates (or bioactive metabolites). But this must be tempered with the possibility that a drug's formulation (e.g., with stearic acid) may reduce absorption and thereby increase excretion of the parent drug. Drug interactions with other chemicals and the user's physiological/disease condition can also dramatically reduce uptake and thereby enhance excretion (or expulsion, e.g., through vomiting) of the parent drug; for example, chelation, alteration in gastrointestinal mobility, or alteration of gastric pH (e.g., chelation of tetracycline by dairy products or of fluoroquinolones by divalent cations). Data also required for these excreted compounds include their expected relative partitioning between STW solids and aqueous effluent, and their expected biodegradation efficiency within the STW. Data for other potential transformation pathways (e.g., photolysis) would be useful. While knowing the effective dosage for a drug's therapeutic target-effects in humans (and domestic animals) is useful, non-target, non-therapeutic effects (e.g., side effects) must also be considered, and these can occur at much lower doses (below low- to sub-µg/kg level -- or ppb-ppt). Knowledge of potential aquatic effects would be most useful for directing final selections.
Aquatic organisms -- captive to continual, life-cycle chemical exposures: Any chemical introduced to the aquatic domain can lead to continual exposure for aquatic organisms. Chemicals that are continually infused to the aquatic environment essentially become "persistent" pollutants even if their half-lives are short. Their supplies are continually replenished and this leads to life-long multi-generational exposures for aquatic organisms. Continual exposure can add an extra dimension to the exposure dose required for eliciting an effect (by reducing it even further). Of aquatic life, fish in particular are sufficiently mobile (as a result of natural behavior or avoidance responses) as to avoid continuous exposures to toxicants. This ability can obviate the course of what would ordinarily be captive, continuous exposure to any perpetual situation, leading to reduced, intermittent exposures. But, depending on the mechanism of action for a toxicant, such transient, intermittent exposure can still have comparable - and persistent - effects. For example, the estrogenic response (vitellogenin production from exposure of fathead minnows to estradiol) from intermittent exposure exceeds that which occurs from an equivalent time-integrated dose from a continuous exposure (38). Therefore, for at least certain toxicants, intermittent exposure (resulting either from episodic/fluctuating concentrations or from contact avoidance) does not necessarily lessen the degree or duration of the effect.
Effects -- from acute to subtle: By their nature, pharmaceuticals are mostly designed to be highly bioactive -- many exquisitely so. Their intended biological targets (e.g., receptors) are frequently extremely specific and sensitive; "receptors" essentially serve as switches that upon receiving an incoming "command" then generate an outgoing signal, the value of which is determined by whether the command is blockage by an antagonist or facilitation by an agonist. Furthermore, both the target receptors and intended effects can be very different from those of currently regulated pollutants (e.g., industrial chemicals and agrochemicals). Unintended, unexpected effects (e.g., adverse side effects; idiosyncratic drug reactions) can be caused by previously unrecognized drug-receptor interactions, previously unidentified receptors, previously unrecognized inter-drug interactions (e.g., via efflux-pump blockers/inducers or cytochrome P450 inducers/inhibitors), and by a broad diversity in drug-metabolizing/transport phenotypes (genetic polymorphisms). Many accomplish their therapeutic effects by mechanisms that have yet to be elucidated; and mechanisms of action can change depending on dose (2). Finally, drugs often tend to receive market approval based on their short-term effectiveness against "surrogate endpoints" (e.g., blood pressure), not on the basis of long-term efficacy against the target malady (e.g., heart disease).
While our understanding of the spectrum of these variables in humans is not fully understood, they are poorly characterized in aquatic organisms. Effects on non-target organisms (e.g., wildlife - not laboratory organisms) are almost completely uninvestigated. Furthermore, the important point, which is totally obscure to current knowledge, is the significance that these variables might have at the population level (effects on individuals being important only as far as an ultimate impact on the population is concerned). Extremely little is known about the effects of these substances on non-target organisms, many of which have different metabolic pathways and different potential receptors. Effects on aquatic (non-target) species could range from easily detectable acute symptomology to unnoticeable subtle change. What little that is known serves to show that rather low concentrations at least have the potential to exert substantive effects on aquatic life. A few PPCPs are already known to elicit profound effects on aquatic life at very low concentrations. For example, selective serotonin re-uptake inhibitor antidepressants (SSRIs) such as fluoxetine and fluvoxamine can induce reproductive behavior in certain shellfish at 10-10 M (ca. 30 ppt); see the extensive discussion of SSRIs in the chapter in this book by Fong (39). Universal regulatory pathways, such as the control of intracellular Ca2+ concentration for activating sperm, could be affected by any of the many calcium channel blockers. For example, pimozide and penfluridol inhibit sperm activation in sea squirts (40); this perhaps is not surprising since infertility in both human sexes has been linked to numerous drugs (ca. over 150) over the years (including members of chemotherapeutics, anti-psychotics, beta-blockers, antimicrobials, etc., and of course, a wide range of hormones). Male infertility, however, is not commonly screened for using lab animal models. Nifedipine (another calcium channel blocker) has been known to impair the ability of sperm to fertilize (reversible effect). Whether any of these compounds occur in the environment (structural stability might be a major limitation) is not known.
While only very sparse data have been generated so far on the subtle ways in which drugs could affect aquatic life, this line of concern is bolstered and illustrated by the substantially more data that is available for pesticides and other "conventional" pollutants. For example, brief exposure of salmon to 1 ppb of certain carbamate or organophosphorus insecticides is known to affect signaling pathways (via olfactory disruption and pheromone detection), leading to alteration in homing behavior (with obvious implications for predation, feeding, mating, and olfactory imprinting) (e.g., 41,42).
Another consideration is that of additive effects. For therapeutic classes comprising multiple members (SSRIs, calcium channel blockers, and anticholinergics being three examples), even if the individual concentrations of each drug were low in the aquatic environment, the combined concentrations of all members sharing a common mode of action could prove significant (i.e., summation of "toxic units"; see 43). This consideration could be particularly important for those drugs having low therapeutic indices (margins of safety, or therapeutic "windows"). As an example of additive effects, anticholinergics are known for their potential to create an insidious additive "anticholinergic load" problem for elderly patients simply because this broad class of drugs comprises about 20 distinct, otherwise unrelated, therapeutic classes - the combination of multiple drugs as opposed to individual members is problematic (44). Analogous issues can be drawn for the aquatic environment.
Finally, the issue of sub-therapeutic exposures is only a part of the overall concern being that continual, life-long exposure to trace levels of any substance is a relatively unexplored domain of human or environmental toxicology. This is one reason that trends data are important. Even where PPCP concentrations are currently low, no data sets exist for revealing long-term trends. For any particular PPCP at a given geographic location, whether its concentration over time is decreasing, remaining constant (steady state), or increasing could dictate whether any action needs to be contemplated to control its introduction to the environment.
Could subtle effects accrete to unnoticed change? A major issue yet to be addressed by ecotoxicological science (and of potentially critical importance for risk assessment) is the impact on non-target species (such as aquatic biota) of stressors eliciting effects (perhaps via low but continual concentrations) sufficiently subtle as to go unnoticed in real time - but whose cumulative impacts eventually yield recognizable outcomes having no obvious cause (1). How would the impact of such cumulative, unrecognizable subtle effects compare with the overt, acute impacts of known toxicants? Unequivocal environmental impacts or change, by their obvious nature, are easily detected, and measures to mitigate untoward outcomes can be engineered and implemented. Change that takes a surreptitious tack, however, grows unnoticed and can result in outcomes not recognized as such -- seeming to arise from nowhere or perhaps mistaken as a natural result of adaptation or evolution within the interconnectedness of nature. While any immediate biological actions on non-target species (unintentional effects, especially in the aquatic habitat) might be imperceptible, we must ask whether PPCPs nonetheless have the potential to lead to adverse impacts as a result of subtle effects ("silent toxicity") from low, ppb-ppt concentrations (µg-ng/L) that impart latent damage. This problem would be exacerbated if the resultant outward change were rationalized simply as resulting from natural evolutionary change or variance and no remediative messages implemented.
This type of ecological impact could persist indefinitely, with no mitigative measures ever put in place. Antineoplastics, SSRIs, and calcium-channel blockers are three drug classes that are known to have the potential for long-term aquatic effects; neuroteratogenic anti-epileptics have the potential for adverse aquatic effects. Subtle, unnoticed effects could accumulate over time until any additional incremental burden imposed by a new, unrelated stressor could possibly trigger sudden collapse of a particular function or behavior across a population. While PPCP residues by themselves might well be below even the thresholds required for subtle, latent effects, their presence may add to the overall aggregate of other, unrelated toxicants - leading to an unacceptable overall burden. An associated risk is the loss of genetic diversity caused by selection against those organisms with insufficient xenobiotic detoxification pathways.
In a range of articles on behavioral and neuro toxicology (see 45,46, and references cited therein), a cogent argument is presented that neurobehavioral effects (with humans being the subjects of concern) are particularly apt to display with such subtlety that they can escape detection by conventional quantitative approaches. Many of these effects are manifested as organisms age and mature -- through subtle alterations in functions such as capacity to learn, motor-sensory performance, and reproduction, coupled with a gradual age-dependent decline in their requisite compensating mechanisms. This makes detection of these slow, degenerative effects, and their distinction from the "natural aging process", extremely difficult.
Weiss makes the argument that truly "abnormal" behavior has the potential to masquerade as seemingly normal deviation within a natural statistical variation. Change can occur so slowly that it appears to result from natural events - with no reason to presume artificial causation. Connections of cause and ultimate effect are difficult to draw, in part because of the ambiguous and subjective nature of the effects, but especially when they are confounded as aggregations of numerous, unrelated interactions.
Effects that are sufficiently subtle that they are undetectable or unnoticed present a challenge to risk assessment (especially ecological). Deviations in behavior within statistical variations are particularly problematic. Advances are required in development and implementation of new aquatic toxicity tests to better ensure that such effects can be detected. While certain fundamental biochemical pathways (e.g., various developmental signaling pathways) are evolutionarily conserved - being the same across disparate species - toxicokinetic differences are still substantial between humans and animals. Just as animal models are frequently called into question for their relevance to human health, mammalian toxicity data (e.g., for PPCPs) may not necessarily be transferrable to aquatic organisms, especially with respect to behavior. The diversity and importance of animal behavior is extremely complex (the focus of ethology, and can be explored in depth at: http://www.york.biosis.org/zrdocs/zoolinfo/behav.htm [post-publication note -- URL changed to: http://www.biosis.org/zrdocs/zoolinfo/behav.htm]). Even the correlation between mammalian toxicity (on the basis of dose) and aquatic toxicity (on the basis of concentration) for conventional tests is non-existent for well-known PBTs (47). It should also be emphasized that the priorities for selecting PPCPs for toxicological evaluation can not be based solely on their relative rankings with respect to environmental concentrations because drugs can dramatically vary with respect to the concentrations at which they impart effects -- sometimes by orders of magnitude.
Data supporting this subtle-effects scenario (posed by Daughton and Ternes ) was recently published (32). Thresholds for estrogenic responses (vitellogenin production in male fish) are shown to be much lower for real-world chronic exposure (i.e., for wild fish) than for short-time study exposures (i.e., for caged fish). Responses are a function of not just the dose, and the timing of dose, but also duration of exposure. Response thresholds (no-effect concentrations) are continually reduced as exposure times increase. Moreover, no overt changes are noted in any fish (e.g., condition factors, survival, growth) regardless of the STW effluent exposure concentration or duration -- only the less obvious vitellogenin concentrations were found to be affected. Longer exposures led to exponentially higher vitellogenin concentrations with concomitantly lower response thresholds.
Efflux pumps: A foundation for aquatic health in polluted waters -- Efflux pump inhibitors: Health risks to aquatic biota? Over the last decade, the identification and characterization of membrane-based active transport systems that "eject" or "pump" toxicants from inside cells have revealed a number of important roles for these systems. These protein-based ejection systems are often referred to as "efflux pumps" or "multi-drug transporters", the best characterized being the P-glycoprotein-like (Pgp) transporter systems (see discussion in Chapter by Epel and Smital ; and in Bard ). First spotlighted by Daughton and Ternes (1) for their potential significance with respect to PPCPs in the aquatic realm, it is becoming evident that these drug (or multi-xenobiotic) efflux pumps, which only started to be characterized in the 1980s for mammalian cells (esp. tumors) and in bacteria, may be broadly distributed across the entire biotic spectrum -- including aquatic organisms.
These pumps are responsible for actively transporting toxicants (esp. amphiphilic molecules) that gain access to (or are created within) the interior of cells back to the outside of the membrane, thereby preventing the intracellular accumulation of toxicants and bioactive metabolites. "Over expression" of these proteinaceous efflux pumps in tumor cells and pathogenic bacteria is widely recognized as enabling resistance to chemotherapeutic agents and antimicrobials. These systems can also be induced (up-regulated) by one drug, leading to enhanced excretion of other drugs (50). Conversely, any of a diverse array of certain chemicals that are able to inhibit these pumping systems can lead to intracellular penetration by, and accumulation of, toxicants in general, and thereby potentiate adverse effects from extracellular concentrations of toxicants that would otherwise prove benign. Because toxicants cannot be readily removed by an exposed organism whose transport system(s) has been impaired, exposure time to the toxicants is thereby lengthened by intracellular accumulation. Another important metabolic enzyme system that also needs to be considered is that comprising the better-known cytochrome P450 monooxygenase isoforms (see listing at: http://medicine.iupui.edu/flockhart/ ); this system is intimately involved with drug metabolism, and numerous drug-drug interactions occur as a result of its inhibition or induction. Inhibition of any of the many cytochrome P450 monooxygenases (or their helper enzymes) can modulate steroid metabolism and lead to population-wide reproductive effects in certain aquatic invertebrates, such as molluscs (51). Numerous drugs and other toxicants are ligands for the ubiquitous P450 enzyme superfamily. This is the route, for example, by which tributyltin causes imposex in molluscs, thereby serving as an indirect endocrine disruptor; P450-mediated steroid metabolism is also used by crustaceans (51).
While efflux pumps are best known for making it more difficult to get therapeutic doses of drugs to target organs (because of reduced drug absorption/uptake, enhanced drug elimination), much less appreciated is the fact that they provide a fundamental level of protection for aquatic biota -- organisms ranging from fish and crustaceans (see Chapter by Epel and Smital , and discussion in Daughton and Ternes ) to ciliates (52), all of which can suffer from continual exposure to any toxicants present. Now recognized for enabling a significant portion of the growing incidence of antimicrobial resistance among bacteria, these systems play a critical role in protecting cells from toxicants, especially in those environments such as the aquatic realm where filter-feeding organisms in particular suffer continual, maximal exposure to toxicants.
Another consideration is the prevalence in the aquatic environment of chemicals that can induce the expression of efflux pump systems. This would add yet another dimension to these considerations by selectively enriching populations for those individuals with inducible efflux systems and thereby placing the entire population at maximum risk should they eventually be exposed to potent inhibitors of these systems. It is also believed that inducement of efflux pump expression is a means by which resistance to antiseptics can be promoted; this mechanism of resistance is also applicable to antimicrobials that supposedly have non-specific mechanisms of actions (e.g., general membrane disruption/denaturation) and that were previously believed to not be amenable to promoting resistance. Efflux pump induction is a recently recognized new avenue of drug-drug interaction in humans, leading to enhanced excretion of unmetabolized drug (50).
A number of substances (primarily pharmaceuticals) are known to inhibit efflux pumps (via a variety of mechanisms), some of the more potent being verapamil, reserpine, and cyclosporin. These inhibitors are also known as efflux pump "reversal agents", "chemosensitizers", "efflux pump blockers", or "efflux pump inhibitors" (EPIs). Given the fundamental role that efflux pumps may play in the health of biota throughout the aquatic food-web (and therefore entire aquatic ecosystems), attention may need to be devoted prior to environmental assessment for screening new chemicals (esp. drugs) for EPI activity. Little is known, however, about which existing xenobiotics can act as "chemosensitizers" in the aquatic environment -- or their frequency of occurrence in the environment. Evidence shows that samples from polluted surface waters impair the functioning of multi-xenobiotic transport in aquatic organisms more than the analogous samples from less-polluted waters (see references cited in Daughton and Ternes ).
Recent developments show that much more potent, broad-spectrum EPIs (which do not necessarily possess any intrinsic bioactivity) are being designed to resurrect or expand the usefulness of existing drugs (including antimicrobials that are no longer effective) as well as enhance the effectiveness of new drugs. These, currently research-only investigational compounds can show orders of magnitude greater EPI activity than verapamil (see the posters prepared by Microcide Pharmaceuticals, Inc., for the 39th Interscience Conference for Antimicrobial Agents and Chemotherapy: http://www.microcide.com/ICAAC99Posters/ icaac99_posters.html [post-publication note -- Corporate name changed. See new information at: http://www.essentialtherapeutics.com/rnd_pubs.html; http://www.essentialtherapeutics.com/prod_pipeline.html] ). The potential for release of such potential future EPI drugs may require careful assessment for environmental impact. This is but one example of a new and promising route to disease therapy that may require active balancing of the clear benefits for human health against the high potential for aquatic harm.
Questions also arise as to whether the combined additive/synergistic action of numerous EPIs (whether drugs or other xenobiotics) in the aquatic environment could be one (of a number) of possible causes for sudden, unexplained mass die-offs of various species. Could systems in apparent good health possibly collapse simply by exposure to one or a series of EPIs (which by themselves would not prove toxic) -- by potentiating the action of toxicants that were already present? How far-reaching could the action of EPIs in the aquatic realm be? The possibly widespread expression of efflux pumps in aquatic organisms also begs the question as to the relevance of trying to correlate lipid burdens of toxic pollutants with organisms' health (e.g., see43) if these concentrations have little if any bearing on the concentrations that actually effect intracellular exposure. Can fat-accumulated chemicals act as reservoirs of toxicants that gain access to intracellular receptors once efflux pumps are shut down?
Finally, it is worth noting that aquatic toxicity is traditionally measured in the absence of efflux pump inhibitors. This means that EC50 values may represent conservative, upper estimations for a given toxicant being that any EPIs that may be present will effectively increase the actual dosage that enters exposed cells -- thereby lowering the EC50.
Are the current approaches to risk assessment comprehensive? Questions can be raised as to whether the approaches to environmental risk assessments and epidemiological studies sufficiently consider the "universe" of toxic substances involved in exposure or whether the focus on conventional "priority pollutants" gives a narrow perspective. By ignoring exposure to chemicals with significant bioactivity (such as pharmaceuticals), can the usefulness of such studies be questioned? What might be the significance of not factoring individual risks from previously unidentified pollutants in discussions of overall risk? This question is addressed in the ecological risk rankings published by the EPA's Science Advisory Board (53). Gross within-class differences with respect to aquatic effects possibly make the approach of assessing eco risk on a class-by-class basis untenable. For example, some SSRIs are extremely potent with regard to shellfish reproductive behavior while others have almost no effect. By not factoring in multi-chemical effects (such as from efflux pump inhibition), potentially large effects could be overlooked.
Why the concern over exposures that are well below therapeutic doses? It is important to recognize that recommended therapeutic dosages can be higher than the therapeutic dose actually required (a dual consequence of not adjusting doses downward from those set in clinical trials and of not individualizing therapy). Non-target, unintended effects cannot be discounted at sub-therapeutic dosages (unintended effects can occur at lower concentrations than do therapeutic effects, especially during long-term maintenance therapy). Additive/synergistic effects from simultaneous consumption of numerous drugs are essentially unknown (intake of multiple drugs adds to the complexity of additional burden for a patient already taking medications with low therapeutic indexes). Finally, continual, life-long exposure to trace levels is an unexplored domain of toxicology.
Relevant to these points regarding exposure to low environmental concentrations is the very important fact that trends data are lacking. Even where PPCP concentrations are currently low, no data sets exist for revealing long-term trends. For any particular PPCP at a given geographic location, is its concentration over time decreasing, remaining constant (steady state), or increasing?
Continually expanding uses and consumption of individual PPCPs points to concentrations of existing drugs increasing in addition to adding to the burden of discrete drug types in the environment. Will those PPCPs with longer half-lives tend to accumulate? Drug use can be expected to continue to rise due to a confluence of drivers: increased per capita consumption, expanding population, expanding potential markets (partly due to mainstream advertising/marketing), expiration of patents, increased use of physician samples, new target age groups, and new uses for existing drugs. Old therapeutics are being used not just for additional clinical conditions (those for which they were not originally developed) but also for non-disease states -- for example, medical manipulation or alteration of personality traits and satisfaction of certain social needs -- referred to as "cosmetic pharmacology."
Finally, there are potential, intangible benefits in being proactive versus reactive -- the Precautionary Principle - the principle of precautionary action that redistributes the burden of proof ("reverse onus") because the science required for truly and fully assessing risks lags far behind the requisite supporting science
Future considerations -- Drug sewage monitoring as a tool: A significant aspect of our society that illustrates the potential impact of our lives on the environment -- as well as the potential impact of our activities on ourselves and others -- is the widespread use of illicit drugs. The potential for the occurrence of illicit drugs (and drugs of abuse) in sewage influent presents a unique and powerful opportunity for a new tool to raise the public's consciousness of drug use -- down to the local community level. By monitoring sewage influent for illicit drugs (and perhaps by back calculating with pharmacokinetic assumptions), estimates of drug use at the community level could be ascertained, while preserving the anonymity of all individuals. Such data have the potential for not only lending a new perspective to illicit drug use, but also for providing a new tool to enhance society's knowledge and perspective regarding the manufacture, trade, and use of these substances. This idea is expanded upon in the last chapter of this book (23).
Summary of Under-Addressed Issues Regarding Environmental Ecotoxicology
In an examination of the world's disparate and disjointed literature involving pharmaceuticals in the environment, Daughton and Ternes (1) highlight a number of issues not frequently encountered in discussions of environmental toxicology. These and others may deserve further attention and debate.
Conclusions and Recommendations
The future for research on PPCPs in the environment: The poorly characterized ramifications of PPCPs in the environment (occurrence, trends, fate, transport, effects), or their relative importance compared with conventional chemical pollutants such as PBTs, warrants a more precautionary view on their environmental disposition. A portion of the effort that continues to be invested in elucidating the environmental transformation and fate of POPs/PBTs might more profitably be redirected to PPCPs.
A major task of the environmental science community might be to examine each therapeutic or consumer-use class of PPCPs (or those grouped according to purported modes of action) and rule-in or rule-out possible deleterious environmental effects on the basis of those that are known to occur in the environment at significant individual or combined concentrations, or those whose concentrations are trending upward. One area that could be pursued immediately is an exhaustive search of the literature for unintended, unexpected effects on non-target species, especially aquatic. Particular scrutiny should be given to the efflux-pump-inhibitory (and efflux pump-inducing) activity of existing and all new drugs with regard to aquatic organisms. Future designs of STWs should perhaps focus on the most difficult-to-remove, toxicologically significant compounds so as to maximize the removal of all anthropogenic pollutants (even those not currently recognized). This work will involve simultaneous contributions from both exposure and effects scientists working in parallel and in sequence (an iterative process).
Near-term objectives to consider for minimizing introduction of drugs to the environment or to minimize potential environmental effects:
The implementation of many of these considerations might contribute not just to minimizing the burden of PPCPs in the environment, but might also benefit end-users and patients by way of minimizing their exposure to PPCPs via end-use and the concomitant universe of side-effects and bacterial resistance development. By holding as a major objective the maximization of the usefulness of toxicological studies for risk assessments, such assessments would become more useful to the good of the public and the environment. In the final analysis, we need to ask whether the current defacto, no-cost approach for dealing with the introduction of PPCPs to the environment - namely dilution by receiving waters - is the best, or if one or more proactive approaches should be implemented.
ACKNOWLEDGMENTS: The author thanks the following people for their valuable time in providing helpful review of both technical and policy aspects of this manuscript: Octavia Conerly (U.S. EPA Office of Water), Michael Firestone (U.S. EPA Office of Pollution, Prevention, and Toxics), Jerry Blancato (U.S. EPA National Exposure Research Laboratory), Thomas Ternes (ESWE, Germany), and ACS's anonymous reviewers, all of whom contributed to improving the quality of the manuscript.
NOTICE: The U.S. Environmental Protection Agency (EPA), through its Office of Research and Development (ORD), funded and performed the research described. This manuscript has been subjected to the EPA's peer and administrative review and has been approved for publication. Mention of trade names or commercial products does not constitute endorsement or recommendation by EPA for use.
NOTE: This chapter is reprinted with permission from the American Chemical Society. It can be cited as follows:
Daughton, C.G. "Pharmaceuticals in the Environment: Overarching Issues and Overview," in Pharmaceuticals and Personal Care Products in the Environment: Scientific and Regulatory Issues, Daughton, C.G. and Jones-Lepp, T. (eds.), Symposium Series 791; American Chemical Society: Washington, D.C., 2001, pp. 2-38.
[Posted (abstracted/excerpted) with permission from the American Chemical Society: Daughton, C.G.; Jones-Lepp, T.L. (eds.) Pharmaceuticals and Personal Care Products in the Environment: Scientific and Regulatory Issues, Symposium Series 791; American Chemical Society: Washington, D.C., 2001.
Copyright 2001 American Chemical Society]