More of the TCE Story Missed by the LA Times

There are a lot of apprehensive people living over TCE vapor plumes, or who have been drinking water at one time was contaminated with TCE, and they need some reliable information about the potential health risks involved. What they get from the LA Times is this:

Following four years of study, senior EPA scientists came to an alarming conclusion: The solvent, trichloroethylene, or TCE, was as much as 40 times more likely to cause cancer than the EPA had previously believed.

In the blink of an eye, this factoid has spread nearly everywhere, Wikipedia included.

How real is this? Let’s take a stab at figuring out where it might come from.

A little bit of background first: one of the tools used in risk assessment is a cancer potency factor (or slope factor). The slope factor is used to calculate the increased cancer risk from exposure to a chemical. Essentially, the cancer risk associated with a scenario is calculated by multiplying the slope factor by the level of chemical exposure.

In 1985, EPA estimated an oral slope factor for TCE of 0.011 kg-day/mg, based on the tumor incidence in an animal bioassay. This value was removed from IRIS in 1989, when EPA’s Science Advisory Board criticized EPA’s classification of TCE as a probable human carcinogen. In its 2001 reassessment, EPA presented a range of slope factors from 0.02 to 0.4 kg-day/mg. Some people (and most likely not EPA scientists, or at least not the ones who really understand this stuff) simply took the high-end slope factor from the 2001 assessment, divided it by the older number, and concluded EPA had found that TCE was 40-fold more carcinogenic in the intervening time between the original assessment in 1985 and the 2001 reassessment.

It’s probably not that simple.

In 1985, the slope factor for TCE was developed from liver tumors in mice, based on indirect evidence (here and here) – I can’t lay hands right now on the primary reference, EPA’s 1985 health assessment (Health Assessment Document for Trichloroethylene, EPA/600/8-82/006F)*. Also, I’m assuming that the 1976 NCI cancer bioassay was the one used by EPA based on chronology – the repeat TCE cancer bioassay using epichlorohydrin-free TCE was not published until 1990. The doses in mice were extrapolated to human body size, then a linear extrapolation model was used to calculate the slope factor that provided an upper bound of the risk associated with exposure to TCE.** The process is described in more detail in EPA’s old 1986 cancer risk assessment guidelines.

In 2001, the value on the upper end of the slope factor range was estimated from relative risk ratio for non-Hodgkins lymphoma observed in an epidemiological study of drinking water exposures in 75 New Jersey communities. The relative risk ratio of 1.4 was combined with the U.S. background incidence of non-Hodgkins lymphoma and the average TCE concentration in drinking water to calculate a unit risk factor (0.00001 per ug/L) which was then converted to a slope factor using the standard adult body weight (70 kg) and drinking water ingestion (2 L/day) assumptions.

One purpose for slope factors is to be able to compare the potencies of different carcinogens, but the comparison isn’t terribly meaningful without these qualifiers attached to them. The more important point, which the Times didn’t bring out, is that EPA’s reassessment included evaluation of the risks potentially associated with non-Hodgkins lymphoma and leukemia. Concerns appear about these effects appear in part to have emerged from the potential drinking water exposures in Woburn, MA (of “A Civil Action” fame). The early epidemiological studies (published after EPA’s 1985 health assessment) apparently were considered to be inconclusive. However, the literature shows more epidemiological studies being conducted in the 1990s of populations potentially exposed to TCE (among other contaminants) in drinking water. So, the story isn’t so much a matter that TCE is a more potent carcinogen than previously believed (that may or may not be the case), but that further research has potentially identified some additional health effects associated with TCE exposure.

I’m disappointed with the LA Times TCE series for a couple of reasons. First, instead of the “40-times-more-carcinogenic”, lazy-man’s hazard evaluation; they would have provided more of a service to their readers by turning to an authoritative review of TCE, such as from the NTP Report of Carcinogens, and producing a readable summary of it (or perhaps even talking with someone at the UCLA School of Public Health, and getting a quick TCE tutorial). Second, for not conducting some real investigative journalism where it was really needed; you get hints and assertions about the problems with the TCE review process, but never any real evidence or narrative about what actually happened (that’s a topic for another post – this one’s long enough already. I’m not quite done with the Times series yet. . .). Producing those sorts of stories is why we really need the mainstream media, because they have the resources to investigate them. But, I feel as if the Times let us down in this case.

----------------------
*EPA has a lot of older documents online here, including the 1987 addendum; but not the 1985 report.

**For slopes estimated using this 1986 methodology (still in use for a lot of chemicals), EPA includes the qualifier that the procedure produces, “. . . a plausible upper limit to the risk that is consistent with some proposed mechanisms of carcinogenesis. Such an estimate, however, does not necessarily give a realistic prediction of the risk. The true value of the risk is unknown, and may be as low as zero.” I’ve always thought this phrase should be inserted as a footer on each page of a risk assessment report.