CDC Finds Low Levels of Diethylene Glycol in OTC Health Products

While updating information on the victims of the mass poisoning due to Elixir of Sulfanilamide in 1937 (see here for the work in progress), I came across 3 intriguing articles published just this month on the detection of diethylene glycol (DEG) in consumer products. Two of these articles report the discovery of DEG (albeit at tiny levels) or a chemical cousin, ethylene glycol.

In the Journal of Toxicology Journal of Medical Toxicology,* investigators from the CDC quantified the amount of DEG and triethylene glycol (TEG), a related compound, in a “convenience sample” of over-the-counter “health products” imported from Asia. The investigators found detectable, but miniscule, DEG levels in 15 (22%) of 68 samples and TEG levels in 2 (3%). The range of DEG levels, in volume-to-volume units, 0.00007%-0.01%, was at least 810 times less than that found in the DEG-contaminated cough syrup in Panama in 2006 (which was reported at a level of 8.1%). The CDC researchers concluded that “these levels probably do not represent an acute public health threat,” but that “additional research focusing on why DEG is found [at all] in these products…is needed.” Furthermore the minimum amount of DEG necessary for toxicity is unknown.** [See important update/addendum below.]

Then in Molecular Pharmaceutics, investigators at the University of Wisconsin and in South Africa wrote of their discovery of ethylene glycol crystals in a compounded suspension of rifampin. The “exact source” of the ethylene glycol in the antimicrobial is unknown. The authors concluded, “[T]he results of this study show how important it is to ensure that the drug and excipients comply with pharmacopeial or FDA standards.”

Perhaps to show that the surveillance for DEG is top-of-mind at the FDA, scientists at the agency wrote in Applied Spectroscopy (who doesn’t subscribe?)* that “portable Raman spectrometers” reliably detect DEG in pharmaceutical-grade glycerin at a limit of 0.32%.

* By the way, I do believe that reports of research supported by US tax dollars should be freely and readily available to American citizens, without having to pay for the article (and regardless of the policies of the journal in which an article’s published).

** DEG toxicity appears to depend on the production of the toxic metabolite 2-hydroxyethoxyacetic acid (HEAA) through the enzymatic actions of alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ADLH), which are known to vary on the basis of genotype and other factors. Variations in enzyme activity may explain the susceptibility of some individuals to DEG poisoning, while others remain unaffected. Notably the administration of fomepizole, an inhibitor of ADH, is a known antidote to DEG poisoning (along with dialysis).

Update: Melgardt de Villiers, lead author of the Molecular Pharmaceutics article, responds by e-mail that the rifampin-derived ethylene glycol crystals were from (or originally observed in) South Africa.

04/06/11 addendum: A review of the full CDC article (which was supplied gratis by the publisher, Springer, at the presumptive request of the corresponding CDC author) reveals the following pertinent facts:

• TEG was assessed in OTC products on the basis of its detection in DEG-contaminated teething syrup in Nigeria (in 2009).
• The convenience sample consisted of OTC liquid- or ointment-based products imported from Asia and sold locally (in Chamblee, Georgia, a suburb of Atlanta). The products were obtained from 9 different stores selling “Asian medicinal products” in the area. (So we’re not talking about OTC brandname drugs, like liquid Tylenol.)
• The products were classified by the authors as “dietary/herbal supplements, eye drops/ointment, ear drops/ointment, miscellaneous health-related tonics, nasal spray, throat/cough drops, or topical agent,” and had names like “Cough Be Gone,” “First Lady Cough and Cold Syrup,” and “Shenji Royal Jelly in honey base.” 
• Most of the 85 products analyzed came from China (71). Ten originated in Hong Kong; 2 came from Singapore, and 1 each were imported from Taiwan and Mayalsia.
• Only 68 products were “analyzable,” because 17 weren’t water soluble—a requirement for the analytic technique used.
• The median amount of DEG (in volume-to-volume [v/v] units) in the 1995 mass poisoning in Haiti was 14.4%, and that in the 2009 mass poisoning in Nigeria was 19.3% (see the Pathophilia-constructed table below).
• Although information about TEG toxicity is scarce, write the authors, the substance “is considered to have a low order of toxicity.” There are only 2 reports of human poisoning with TEG in the medical literature (see here and here). Both were cases of attempted suicide by, for instance, drinking brake fluid.
• The percentage of TEG (v/v) in the Nigeria mass poisoning was 0.6% (vs the miniscule 0.0012% and 0.0018% found in the 2 products in this study).
• On the basis of data from historical poisonings, including the Elixir of Sulfanilamide incident of 1937, the median or average doses of DEG associated with toxicity or lethality in humans range widely from 14 mg of the substance per kg of body mass (Argentina, 1992) to 1500 mg/kg (Haiti, 1995).
• Given the very low levels of DEG detected in the study (and only in some products), the authors speculate that it “may simply be a minor impurity created during the manufacturing processes of chemicals used to formulate drugs and consumer products.”

Contamination Source	DEG, v/v %	TEG, v/v %
Imported OTC products, 2009	0.00007–0.01 (n = 15 of 68)	0.0012, 0.0018 (n = 2 of 68)
Haiti, 1995 (median)	14.4	—
Panama, 2006 (mean)	8.1	—
Nigeria, 2009 (mean)	19.3	0.6