April 3, 2007

Environmental Effects of Human Pharmaceuticals

By Sumpter, John Phillip

In the last few years, it has been demonstrated that many human pharmaceuticals are present in the aquatic environment, albeit at very low concentrations. Nevertheless, in at least one case, that of ethinyl estradiol, it is very likely that in some locations ethinyl estradiol is adversely affecting fish through its "feminization" of males. Another pharmaceutical, diclofenac, has caused the deaths of millions of vultures in Southeast Asia through its use in veterinary medicine. Those examples have highlighted our lack of understanding of fate, behavior, and effects of pharmaceuticals once they reach the environment. It is unclear presently whether these are isolated examples or whether there is a more general problem. If wildlife are to be protected from the effects of current and new pharmaceuticals, then our knowledge and understanding of this issue need to improve substantially. The recent European Medicines Agency guidelines covering the environmental risk assessment of human pharmaceuticals are a step in the right direction, but a more sophisticated approach, rather than a "one-fits-all" solution, is probably needed. Recent collaboration between industry scientists and academic ecotoxicologists is undoubtedly an encouraging development that is likely to markedly increase the quality, and hence value, of research undertaken in this area.

Key Words


Ethinyl estradiol;


Adverse effects;

Aquatic environment


There is now no doubt that many human pharmaceuticals (and probably also their metabolites) are present in effluents of sewage treatment works (STW; see article in this issue by Williams and Cook). Concentrations generally seem to be in the nanogramper-liter range, with perhaps a few present at very low microgram-per-liter concentrations. There will be dilution of the STW effluent in the receiving water (usually a river) and further degradation, such that aquatic organisms will be exposed to very low concentrations. Is it conceivable that such low concentrations could cause effects in exposed wildlife (such as fish), and if they did, what effects are likely, and what might be the consequences of any of these effects? Only when we have answers to these questions will it be possible to gauge how serious, or not, the issue of pharmaceuticals in the environment actually is. Currently, it is my opinion that, for most pharmaceuticals, we are a long way from having the answers we require, and hence it is not possible to conduct meaningful environmental risk assessments for most pharmaceuticals in use today.


There are a variety of reasons to think that it is at least theoretically possible that human pharmaceuticals could adversely affect wildlife, and an equally long list of reasons exists to think that they are unlikely to do so. Some of the key reasons are given in Table 1.

There are currently only two examples known for which there is strong evidence to suggest that a human pharmaceutical has harmed wildlife. One is the "feminizing" effect that ethinyl estradiol (EE2) has on fish, and the other is the death of vultures because of diclofenac poisoning. EE2 originates from the use of the contraceptive "pill." It is one of a number of steroid estrogens, both natural and synthetic, found in STW effluents in many developed countries (1). Fish exposed to these estrogenic effluents have elevated plasma vitellogenin concentrations (vitellogenin is a biomarker of estrogen exposure), and the males can be intersex (2). In at least some countries, such as the United Kingdom, EE2 appears to contribute significantly to the total estrogenic activity of effluents and rivers (3). Fish are exquisitely sensitive to continuous exposure to EE2; the midpoint of the dose-response curve for vitellogenin induction occurs at less than 1 ng/L. Full life- cycle assessments done in the laboratory have shown that at an EE2 concentration of only a few nanograms per liter, male fish are feminized to varying degrees and are unable to reproduce (4,5). A large field study in Canada, in which an entire lake was treated with EE2, has shown that EE2 has dramatic effects on fish at a concentration of approximately 5 ng/L (6). These results, produced by both academic and industry scientists, are indisputable.

The second is an even more dramatic example of an adverse environmental effect of a pharmaceutical. Many millions of vultures have unnaturally died across the Indian subcontinent in the past 10 years. Populations of three species of vulture have declined by more than 95%, and all three are now considered critically endangered (ie, likely to become extinct in the wild soon). The birds died of acute renal failure caused by the pharmaceutical diclofenac. They ingested the diclofenac through feeding on the carcasses of livestock (eg, cows) that had been treated previously with the nonsteroidal anti-inflammatory drug diclofenac (7). Direct oral exposure of captive vultures and feeding vultures diclofenactreated livestock confirmed that diclofenac was responsible for the catastrophic decline of vulture populations. Modeling studies show that only a small proportion (less than 1%) of ungulate carcasses needed to be contaminated with diclofenac to have caused the precipitous population declines (8).

The case study of diclofenac in particular demonstrates a number of lessons. One is that pharmaceuticals can reach the environment through unexpected routes. Admittedly, in this particular case diclofenac was used as a veterinary pharmaceutical (although it is still also widely prescribed as a human pharmaceutical), but this does not negate the point. Another is that predicting which animals will receive exposure to pharmaceuticals is not easy. In retrospect, it is obvious that fish will receive low-level exposure to EE2 in some situations, but I do not suppose many of us would have placed vultures near or at the top of our list of wild animals likely to be exposed to pharmaceuticals. Other lessons from the case of diclofenac and vultures include the probability that some species (or groups of species) may be particularly sensitive to a particular pharmaceutical (it is difficult to believe that the dose of diclofenac ingested by the vultures was that high), and that an occasional side effect of a drug in humans may be the dominant effect in other species. All these lessons serve to illustrate that there is still a lot to learn about the (possible) effects of pharmaceuticals on wildlife. More lessons will undoubtedly emerge as our knowledge increases.

It is unclear presently whether these two examples will prove to be atypical-perhaps even the only examples of human pharmaceuticals adversely affecting wildlife-or whether more, perhaps even many more, examples of human pharmaceuticals adversely affecting wildlife will be discovered.


As part of our research into the (possible) adverse effects of human pharmaceuticals on aquatic organisms, we have been studying clofibric acid, an "old" cardiovascular drug now usually replaced by newer fibrates. Clofibric acid has been widely reported to be present in the aquatic environment (reviewed in Ref. 9), and hence in some situations fish will receive continuous low-level exposure to this drug. In mammals, fibrates alter lipoprotein metabolism, so based on the read-across hypothesis developed by Huggett and colleagues (10) we investigated lipid metabolism in fish exposed to clofibric acid. In a series of experiments, we were unable to demonstrate consistent effects of clofibric acid on either plasma cholesterol or plasma triglyceride concentrations (T. J. Runnails, D. N. Hala, and J. P. Sumpter, unpublished results). This was despite designing the experiments in such a way that our chances of detecting any changes were considerably increased above that of a typical ecotoxicology test. Thus, the results indicate that clofibric acid does not affect lipoprotein metabolism in fish. However, is this the right conclusion? Careful scrutiny of the clinical data for this drug shows that clofibric acid produces, at most, a 20% fall in cholesterol and triglyceride levels in patients, and this effect is only observed in patients with abnormally high levels. In normal (not hyperlipidemic) people, clofibric acid has no discernible effect on plasma cholesterol and triglyceride levels. Therefore, perhaps it was unsurprising that we failed to observe a reproducible effect. Further, it seems quite possible that clofibric acid does not affect lipoprotein metabolism in "normal" fish, and hence no amount of trying will produce a reproducible effect.

There may well be an important message here. Many human pharmaceuticals produce relatively small effects (although these may clinically be very important). Even if these did occur in wildlife, they will be very difficult, if not impossible, to detect with confidence in ecotoxicology tests. Consider β-blockers as a "typical" example. These are prescribed to patients with high blood pressure and other cardiovascular problems. They are beneficial through their ability to cause a decrease in heart rate and heart contractility. Because of their widespread use, it is unsurprising that many different β-blockers have been detected in the aquatic environment. Fish have β-adrenergic \receptors just like those of humans, so presumably could respond to a β- blocker in the same way. But even if they did, there are many obstacles to demonstrating this.

First, no current ecotoxicology test (let alone a fully validated one) has been developed that assesses possible cardiovascular effects. Measuring heart rate and blood pressure in a fish is not a trivial undertaking. Even if it could be done reliably and routinely, could, say, a 20% change in one or both of these parameters be detected? How many experimental animals (fish) would be required to produce an experimental design capable of detecting changes of that magnitude (which is at the upper end of the changes induced in patients taking β-blockers)? Ethical considerations mean that scientists should be attempting to reduce the number of animals used in research (11), not increasing the number in the (possibly unsuccessful) hope of detecting a small change in one or more physiological parameters. Further, even if a small effect can be detected, it is important to know whether it is reproducible, whether the effect is adverse, and what the consequences (to the animal) of the effect are. It will not be an easy task to provide answers to these questions.


Recently, the European Medicines Agency (EMEA) published its "Guideline on the Environmental Risk Assessment of Medicinal Products for Human Use" (12). This document is undoubtedly a step in the right direction and is a significant improvement on the previous situation. It recommends ecotoxicity tests be conducted on three aquatic species: one plant, one invertebrate, and one vertebrate. These would be an algal growth inhibition test, a daphnia reproduction test, and a fish early life-stage test. It is unclear presently whether the data set produced will provide the information needed to fully protect the environment. The proposed suite of tests would not, in my opinion, have detected the environmental effects of either EE2 on fish or diclofenac on vultures. It remains to be seen whether other false negatives will occur.

Given the vast amount of research that has been conducted on EE2 and its effects on aquatic wildlife (especially fish) in the last decade, some lessons ought to have been learned that can be applied generally to the issue of the presence of low concentrations of very biologically active chemicals (such as pharmaceuticals) in the environment (see, eg, Ref. 11). As an example of the type of thinking that could prove helpful, let us consider the synthetic progestogens. These are used primarily in the contraceptive pill. Although they have not as yet been reported to be present in the aquatic environment, it seems likely that they are, in low concentrations, because one might expect them to behave in a similar manner to the synthetic estrogen EE2. But, why be concerned about them even if they are present? Well, fish have progesterone receptors, and endogenous progesterones are required for successful spermatogenesis in males and ovulation in females. Various progesterones are also used as reproductive pheromones; some fish can "smell" pheromonal progesterones when they are present in the water at concentrations as low as 1 10^sup -12^ M (about 0.3 ng/ L). Given this information, it is apparent that synthetic progestogens used as human pharmaceuticals could (and only could) adversely affect reproduction of fish. At the very least, they merit serious study, although to date no relevant results have been published. My interpretation of the EMEA guidelines would suggest that even if synthetic progestogens do adversely affect fish reproduction at low concentrations, this is likely to go unnoticed.


Although I realize that it is almost always unwise to speculate, my personal views on the issue of pharmaceuticals in the environment are as follows: I consider it very likely that many, and perhaps most, pharmaceuticals will be present in the environment at concentrations too low (possibly far too low) to cause any effects on fauna and flora. Many pharmaceuticals may not enter aquatic organisms easily, and most of those that do will not bioconcentrate to any significant extent, so internal concentrations will be low. Even if some pharmaceuticals do cause effects to some groups of organisms, in many cases these effects may be very small (perhaps insignificant). Further, there is no reason to think that all effects will be negative (adverse); pharmaceuticals are intended to do good to patients, not harm, so perhaps some will make fish and other aquatic organisms healthier. What is required is an ability to identify those (probably few) pharmaceuticals that could harm wildlife before they do so. We need to be able to predict which pharmaceuticals will do harm, to which organisms, and at which concentrations. If that situation could be reached, then it should be possible to prevent situations such as the disastrous loss of millions of vultures in Asia. That is the challenge.


1. Desbrow C. Routledge EJ, Brighty GC, Sumpter JP, Waldock M. Identification of estrogenic chemicals in STW effluent. I: Chemical fractionation and in vitro biological screening. Environ Sci Technol. 1998;32:1549-1558.

2. Sumpter JP, Johnson AC, Williams RJ, Kortenkamp A, Scholze M. Modeling effects of mixtures of endocrine disrupting chemicals at the river catchment scale. Environ Sci Technol. 2006; 40:5478-5489.

3. Jobling S, Nolan M, Tyler CR, Brighty G, Sumpter JP. Widespread sexual disruption in wild fish. Environ Sci Technol. 1998;32:2498-2506.

4. Lnge R, Hutchinson TH, Croudace CP, et al. Effects of the synthetic estrogen 17α-ethinylestradiol over the life-cycle of the fathead minnow (Pimephales promelas). Environ Toxicol Chem. 2001; 20:1216-1227.

5. Nash JP, Kime DE, Van der Ven LTM, et al. Long-term exposure to environmental concentrations of the pharmaceutical ethinylestradiol causes reproductive failure in fish. Environ Health Perspect. 2004;112:1725-1733.

6. Palace VP, Wautier KG, Evans RE, et al. Biochemical and histological effects in pearl dace (Margariscus margarita) chronically exposed to a synthetic estrogen in a whole lake experiment. Environ Toxicol Chem. 2006;25:1114-1125.

7. Oaks JL, Gilbert M, Virani MZ, et al. Diclofenanc residues as the cause of vulture population decline in Pakistan. Nature 2004;427:630-633.

8. Green RE, Newton I, Shultz S, et al. Diclofenac poisoning as a cause of vulture population declines across the Indian subcontinent. J Appl Ecol. 2004;41:793-800.

9. Pent K, West AA, Caminada D. Ecotoxicology of human pharmaceuticals. Aquatic Toxicol. 2006; 76:122-159.

10. Huggett DB, Benson WH, Chipman K, et al. Role of mammalian data in determining pharmaceutical responses in aquatic species. In: Williams RT, ed. Human Pharmaceuticals: Assessing the Impacts on Aquatic Ecosystems. Pensacola, FL: SETAC Press; 2005:149-181.

11. Hutchinson TH, Barrett S, Busby M, et al. A strategy to reduce the numbers offish used in acute ecotoxicity testing of pharmaceuticals. Environ Toxicol Chem. 2003;22:3031-3036.

12. European Medicines Agency (EMEA). Guidelines on the Environmental Risk Assessment of Medicinal Products for Human Use. Document Ref. EMEA/CHMP/SWP/4447/00; June 13, 2006.

Job Phillip Sumpter,

BSc, PhD, DSc

Professor of Aquatic

Ecotaxicology, Institute for

the Environment,

Brunei University,

Oxbridge, Middlesex, UK

Correspondence Address

Professor John Sumpter, Institute for the Environment, Brunel University, Uxbridge, Middlesex, UB8 3PH, UK (e-mail: [email protected]).

Research in the author's laboratory is supported by the European Union, NERC, DEFRA, AstraZeneca, and Pfizer.

This article was presented at the joint DIA/HESI/SAPS Conference on Environmental Assessment of Human Medicines, held in Stockholm, Sweden, May 22-23, 2006.

John Sumpter has disclosed that he has received grants/research support from AstraZeneca and Pfizer Inc.

Copyright Drug Information Association 2007

(c) 2007 Drug Information Journal. Provided by ProQuest Information and Learning. All rights Reserved.