Comment on Strupp papers on beryllium metal toxicity.

I read with great interest the research articles by Strupp (2011a,b) ‘Beryllium Metal I’ and ‘Beryllium Metal II’ which were funded by the REACH Beryllium Consortium (2008), an industry group led by Brush Wellman Inc., the largest producer of beryllium in North America. I would like to comment on the author’s conclusion that beryllium metal is not a skin sensitizer and to caution readers on the certainty of this conclusion. First, I will address the method used for the skin sensitization test (Strupp, 2011a) and then comment on the discussion of the data that omitted key papers in the published literature to support the finding (Strupp, 2011b). Organization for Economic Cooperation and Development (OECD) Guideline 406 is a guinea pig maximization test that calls for intradermal injection and subsequent topical application of the test agent (OECD, 1992). Using this protocol, Strupp (2011a) reported that none of the test animals became sensitized following exposure to beryllium metal. As noted in OECD Guideline 406, testing mice using objective validated tests such as the mouse ear swelling test (MEST) and the local lymph node assay (LLNA) is also sufficient to designate a substance as a sensitizer. Tinkle et al. (2003) topically applied beryllium sulfate to ear skin of mice and demonstrated that mice had become sensitized as quantified using the LLNA assay. In a second experiment, Tinkle et al. (2003) topically applied beryllium oxide in petrolatum to bare skin on the backs of mice for 1 day and 1 week later challenged the same mice with beryllium sulfate. Ear thickness, as measured by the MEST, was significantly increased at 48 h after challenge, indicating that beryllium oxide particles had induced sensitization when applied topically to the skin. To induce sensitization, Strupp used intradermal injection of metal powder suspended in polyethylene glycol (PEG) though topical application is an effective pathway for poorly soluble beryllium compounds (Tinkle et al., 2003). Metal powder in PEG was also used for delivery of booster and challenge exposures. PEG has near neutral pH in aqueous solution and dissolution of metal is a factor of 10–40 slower than soluble beryllium chloride at neutral pH (Strupp, 2011a). In contrast, previous studies used soluble beryllium sulfate for challenge to ensure that sufficient ions were available to provoke a reaction in sensitized animals (Marx and Burrell, 1973; Zissu et al., 1996; Tinkle et al., 2003). Hence, reasons why Strupp’s data for metal differ from previous studies using beryllium sulfate and oxide may be due to the method of exposure delivery and/or lack of beryllium ions in the challenge exposure. In discussing the results of the skin sensitization test, Strupp (2011b) omitted several key studies (Marx and Burrell, 1973; Tinkle et al., 2003) and discounted the relevance of studies by Curtis (1951) and Zissu et al. (1996) as these latter two studies were deemed to ‘not give a satisfactory answer for metal’. Such incomplete compilation and assessment of the literature do not provide a full perspective on existing knowledge with regard to the skin-sensitizing potential of beryllium compounds, including the metal. Curtis (1951) demonstrated 60 years ago that beryllium salts were skin sensitizers in humans and subsequent animal studies confirmed this observation (Marx and Burrell, 1973; Zissu et al., 1996; Tinkle et al., 2003). Additionally, hypersensitivity reactions can be elicited from sensitized guinea pigs upon challenge with beryllium metal (Zissu et al., 1996). Importantly, Tinkle et al. (2003) demonstrated that topical application of poorly soluble beryllium oxide powder was sufficient to induce skin sensitization in mice. This finding is especially noteworthy as Cummings et al. (2007) reported that when an occupational preventive program, which included protecting skin, was instituted in an oxide machining facility, the