Tackling a grubby reputation

Triclosan’s bug-killing prowess won it a place on the ingredients lists of handwashes, deodorants and toothpastes for many years but a series of safety concerns have tarnished its reputation. Roz Bulleid traces a debate that has been rumbling on for more than a decade

Ten years ago Sainsbury’s proudly launched a range of "Microban" antibacterial products. Chopping boards, bin bags and dish cloths were all to be impregnated with the biocide triclosan in response to "growing public concern over food hygiene".

But such products have quietly slipped off the shelves following criticism by NGOs and governments. Today, triclosan is restricted to the personal care department, appearing in mouthwashes, toothpastes, liquid soaps and hand cleaners. Even here it is under siege with several retailers promising its elimination.

Triclosan is an organochlorine that damages bacterial cell walls and is effective against some moulds and yeasts. It was developed in the 1960s by JR Geigy, now part of Ciba Specialty Chemicals, and is also used as a surgical scrub in hospitals. Although it is out of patent and now made by several companies in India and China, Ciba’s plant in Germany is the main EU production site.

Competition rules apparently prevent the firm from disclosing how much triclosan it sells, but UK consumers are thought to be some of the biggest users, each flushing about one gram down the drain every year.

While this suggests a national use of at least 60 tonnes, it is possible the figure has fallen with the recent product reformulations. Some EU countries have certainly reported a drop but no one seems to have looked at UK trends.

Growing unease
Early concerns about triclosan centred around two issues: its possible contamination with structurally similar dioxins and the possibility that its widespread use might result in bacteria developing resistance to drugs with the same mode of action.

The first concern, raised by the US Environmental Protection Agency in the early 1990s, does not apply to more recent formulations of triclosan, says Gerald Renner, director of science and research at the European Cosmetic, Toiletry and Perfumery Association (Colipa).

But the issue has not entirely vanished: the UK’s Environment Agency for one is worried about laboratory reports that sunlight can transform triclosan into dichlorodibenzo-p-dioxin.

Researchers have also suggested that children brought up in overly clean homes might not be exposed to enough bacteria to develop their immune systems, leading to asthma and other health problems.

The drug-resistance issue emerged in the late 1990s and has been the subject of several studies. These suggest cross-resistance can be engineered in the laboratory but have not found real-life cases.

When the European Commission’s Scientific Committee on Consumer Products investigated the topic at the request of the Norwegian authorities, it concluded there was "no evidence of clinical resistance and cross-resistance occurring from the use of triclosan in cosmetic products."1

But it also noted a lack of evidence for "any extra health benefits from having antibacterial cleansers in ordinary households" and referred to comments from the US Center for Disease Control and Prevention suggesting ordinary soap and warm water were more appropriate for everyday use than antibacterial soaps.

The stirrings of unease in the late 1990s were enough to persuade the cleaning products industry to back away from triclosan. In 1998, members of the European industry association AISE agreed not to increase their use of the substance (ENDS Report 309, pp 12-13 ). Today, none of its members, or those of its UK equivalent the UKCPI, use triclosan.

By 2000, attention was switching to triclosan’s environmental fate. It was found to accumulate in caged fish kept near sewage outfalls in Sweden and in human breast milk, prompting the Swedish Chemicals Inspectorate (Kemi) and other government agencies in the country to call for an end to "unnecessary uses".

Tests show triclosan is toxic to a range of aquatic species, but especially to algae. Concentrations of around 0.26 milligrams per litre are lethal to fish and 0.13mg/l can kill crustaceans.2

Standard biodegradation tests do not work on biocides as they kill the bacteria involved, but triclosan’s declining concentration downstream from sewage treatment works and laboratory-based UV tests suggest it is broken down by sunlight.

Some nonetheless collects in sediments. UK figures are not available but Ciba reports levels of about one microgram per kilogram of dry sediment in the river Rhine.

An early risk analysis by industry carried out along the lines used for official EU assessments, established a ‘predicted no effect concentration’ (PNEC), below which no adverse effects would be expected, of 0.07 micrograms per litre. It concluded that environmental concentrations were well below this level and therefore proved "no demonstrable risk to the environment," says Mr Renner of Colipa.

A more recent study showed triclosan disrupting the thyroid hormones controlling tadpole metamorphosis, raising questions about its effect on these hormones in humans (ENDS Report 384, p 25 ). Ciba questions the findings, saying the study artificially induced metamorphosis so the same effect may not be seen in the wild.

The company is considering what further tests it might carry out, but is confident that thyroid effects in mammals are unlikely given the battery of tests carried out in the US before triclosan could be approved as a pharmaceutical.

Toxicity debate
Between 2000 and 2003, triclosan and methyl-triclosan - a metabolite which is more persistent than triclosan itself - emerged as ubiquitous contaminants, having been discovered in rivers and lakes in Denmark, Sweden, Switzerland and the US. Several more EU governments issued warnings about its use and reinsurance giant Swiss Re alerted manufacturers and retailers to possible business risks (ENDS Report 333, p 31 ).

The furore prompted Asda, B&Q and Marks and Spencer to promise to remove triclosan from products (ENDS Report 345, p 35 ).

It also persuaded the Environment Agency to publish a briefing paper on the substance (ENDS Report 358, pp 10-11 ), one of the first under its newly established chemicals strategy (see box, p 33 ).

The 2004 briefing paper highlighted concerns about triclosan’s environmental fate and called on the industries using it to investigate levels in rivers and sewage sludge. They were also asked to consider its toxicity to sediment-dwelling organisms, the dangers posed by methyl-triclosan and the consequences of triclosan-contaminated sewage sludge spread on fields.

Some of these questions still remain but the paper did spawn an industry working group on the substance chaired by the Cosmetic, Toiletry and Perfumery Association (CTPA) and comprising Ciba, Colgate-Palmolive, GlaxoSmithKline, Procter & Gamble, Unilever and the Agency.

Moving the goalposts
Triclosan’s existing PNEC is calculated, following standard practice, from the most sensitive of the species it has been tested on, the freshwater algae Scenedesmus subspicatus. Triclosan concentrations above 0.7µg/l inhibit the growth of the alga; this level is then divided by ten to provide a safety margin, giving a PNEC of 0.07µg/l.

One of the first pieces of research under the working group’s auspices was a small survey of river levels carried out by the Agency to confirm triclosan’s presence in UK waters.

When 15% of samples showed levels close to or above the PNEC, the industry members of the group responded with proposals to set a new, higher safety limit.

The proposed level of 0.89µg/l is over ten times higher and based on statistical analysis of toxicity levels across eleven species. The companies claim it gives "a more robust and realistic indication of the threshold of effect in the natural environment" because it takes more species into account - albeit not the most sensitive. They also justify it on the basis that Scenedesmus is only inhibited, not killed, at the lowest concentrations, and that some of the triclosan may not be in a biologically available form.

The new PNEC is four and a half times greater than the highest aquatic concentration recorded by the Agency. But this approach does not take into account methyl-triclosan, which the Agency found at concentrations of up to 0.03µg/l or triclosan’s toxicity to sediment-dwelling organisms. Ciba says separate research carried out for the Agency on methyl-triclosan suggests it is less harmful, but it cannot reveal detailed results yet. It cannot disclose the results of studies on sediment either, but says they "further support the responsible use of triclosan."

Still unknown is the fate of triclosan extracted by sewage works. More than 90% disappears during treatment but it is not clear how much of this ends up in sewage sludge and is then spread on land. Michael Bernheim, who represents Ciba in the working group, points to US-based research carried out by Procter & Gamble five years ago. This showed that most of the triclosan biodegraded, leaving only 1.5-4.5% among the remaining sludge. But more recent work by academics in Germany and the US gives the proportion remaining in sludge as 30% and 50% respectively.3The Agency’s working group has not tried to resolve the question as the variation between sewage works would make it hard to address, says Jo Kennedy, who leads the Agency’s triclosan work.

At the moment she is looking at the case for revising the PNEC, and has asked water companies to add triclosan to the suite of substances they measure in sewage plant effluent. The Agency itself has no plans for extra research.

Regulatory loophole
Despite the concerns raised by Scandinavian governments, there has been no EU-level scrutiny of the environmental risks posed by triclosan’s use in toothpaste and liquid soap. The 1976 cosmetic products Directive which regulates such products only addresses human health risks.

The Directive lists triclosan as a permitted preservative at concentrations of up to 0.3%. Higher concentrations are permitted for other uses but the 0.3% limit is likely to be made absolute soon.

Triclosan is also due to undergo assessment under the 1998 biocidal products Directive later this year. But this Directive, which has cut a swathe through the number of biocides on the market, specifically exempts products covered by the cosmetics Directive (ENDS Report 385, pp 49-51 ).

Ciba is also likely to have to submit an environmental assessment for triclosan under the EU’s forthcoming REACH Directive. Although this covers all chemicals, the volume of triclosan sold may mean it is not a priority.

For Gwynne Lyons of WWF, the questions surrounding triclosan’s environmental effect are symptomatic of a broader regulatory problem. At the moment producers are simply asked to prove that a chemical’s risks are minimal. She would like to see a process where risks are weighed against a product’s advantages.

"One would accept a greater risk for something that’s got greater benefits," she says. "But I think we’re obsessed with antibacterials when good old hygiene is adequate for most uses."

Cause célèbre
In the meantime, the CTPA is confident that the efforts of the Agency working group will clear triclosan: "The work that’s been done has shown triclosan is safe at the levels it’s used at the moment, and I suspect the group will be disbanded," says representative Paul Crawford.

He thinks the attention paid to the substance was unjustified; that it simply became a cause célèbre for campaigners because of its presence in so many products.

Dr Bernheim of Ciba is also confident the Agency’s concerns can be addressed and notes that a "very comprehensive set of data is available" which "proves that triclosan is safe for humans and the environment in the recommended uses."

But however conclusive the research on triclosan is, it also highlights the relatively limited work done on similar substances.

A study on liquid hand soap ingredients published by the Danish Environmental Protection Agency last year showed, for instance, that kathon - a biocide used as a preservative in many UK products - is highly toxic to algae and not readily biodegradable.4 The substance, which is sold by Rohm and Haas, is a mixture of methylisothiazolinone and methylchloroisothiazolinone, both of which can induce sensitivity in humans.

There are no data on the effect of long-term exposure on aquatic organisms and only limited data on short-term toxicity. After estimating national usage, the study concludes that environmental levels in rivers are potentially much higher than a provisional PNEC, although the effect would be mitigated well-mixed waters.

The Agency has not looked for kathon in the environment and has no immediate plans to do so - or to address the impact of other antibacterials.

Possible alternatives
The CTPA’s Mr Crawford is a staunch defender of triclosan, arguing it plays a unique role in toothpastes and deodorants that other substances cannot emulate. Any reduction in the range of preservatives available to formulators would also increase customers’ exposure to the substances left on the list, he argues.

A quick survey of supermarket shelves shows triclosan still being used in plenty of products but some manufacturers seem to have found alternatives. GlaxoSmithKline, for instance, puts triclosan in some of its Sensodyne toothpastes, but not in Aquafresh or Macleans. None of the children’s toothpastes examined by ENDS contained triclosan.

In Norway, where possible bans have been mentioned, only Colgate Total toothpaste still contains triclosan, and there does not seem to be any in the main UK deodorant brands. It has also disappeared from the ingredients lists of Cusson’s Carex soap, Reckitt Benckiser’s Dettol soap and Colgate-Palmolive’s Plax mouthwash.

Deciphering exactly what alternatives are being used is difficult because manufacturers do not have to highlight active ingredients in their labelling. Indeed, some products may not have an "active" biocide at all as the industry’s labelling guidelines only say products should disinfect to a safe level, which can often be achieved using substances such as alcohol or vinegar.

When Asda reformulated its antibacterial soap last year, it found that the triclosan could be replaced by the conditioner cocamidopropyl pg-dimonium chloride phosphate, which has a mild biocidal action. It is still investigating alternatives for use in toothpaste.

Manufacturers of "natural" cosmetics often take advantage of the antibacterial properties of substances such as tea tree oil. Marks & Spencer, for instance, sells a tea tree oil soap and uses citric and sorbic acid, rice wine alcohol and orange terpenes in its range of "natural" cleaning products (ENDS Report 362, p 35 ).

While such substances are less likely to be persistent or bioaccumulative than synthetic chemicals, they may still be harmful as recent concerns about tea tree oil’s endocrine disrupting and cross-resistance potential illustrate.5,6

Concern waning?
It is symptomatic that Sainsbury’s did not extend its licence to use triclosan in Microban products when it came up for renewal and has now reformulated its cleaning products.

But the company’s environment department notes public concern about the substance seems to have waned as energy and resource issues increasingly take centre stage in the environment debate. The absence of a major campaign group focused on chemicals and the subject’s complexity may also have played a role, it suggests.

Sainsbury’s also argues there are still some uses such as toothpaste where triclosan is the best option. It says it reviews all available evidence before formulating new products but there are cases where it is better to stick with triclosan than use less well-researched alternatives.

Colipa is of the same opinion: "I’m always wary of substitution - because, in all fairness, there is no zero risk," says Mr Renner.