In the January 2015 issue of the Townsend Letter, we published "The Antimicrobial Activity of Selected Silver Products," authored by Robert Rowen, MD; Dennis Harper; DC; and Richard Robison, PhD. After further editorial review, the Townsend Letter retracts the publication of this article and the conclusions based on the reported study of silver products. The study cites that one commercial formulation, ACS 200 Extra Strength, had a measurably higher kill rate against MRSA (methicillin-resistant Staphylococcus aureus) compared with four other commercial products. Although the laboratory report does not name the other commercial products, unblinded lab data do list the competitor products as pH Structured Silver, Argentyn-23, Oxysilver, and ASAP. Three of the aforementioned products were listed as having insignificant "log reduction" killing capacity and reduced percentage kill rate. Reputable independent testing of Argentyn-23 by a commercial laboratory adhering to "good laboratory practices" demonstrated major log reduction killing capacity and percentage kill rate of MRSA.
Inasmuch as the independent testing of Argentyn-23 by a laboratory following compendial procedures as defined by the US Pharmacopeia and the National Formulary guidelines is completely in disagreement with the testing carried out by Robison's university laboratory, the entire study by Rowen, Harper, and Robison is called into question and fails to meet the requirements for journal publication. The paper is hereby retracted; all references to this article must show that it is officially retracted with the publication of the May 2015 issue of the Townsend Letter.
The testing conducted by Robison's laboratory, not a commercial laboratory, was authorized by Rowen and Harper. Robison has previously done laboratory testing for the manufacturer of ACS 200, Results RNA. Robison's analyses for Results RNA include "Antimicrobial Activity of ACS 200 Using Candida Albicans" reported in September 2008 and August 2009 and "Antimicrobial Activity of ACS 200 Using Borrelia Burgdorferi" reported in September 2009. Natural Immunogenics Corp., the manufacturer of Argentyn 23, was asked by Rowen to participate in a "blinded" study of silver products. Natural Immunogenics declined to participate in the study when it was informed that Robison would be conducting the analysis and that the protocol for testing would not be disclosed. Nevertheless, Rowen and Harper proceeded to test Argentyn 23 without its authorization.
The testing of the five commercial silver products by Robison had serious laboratory methodology flaws. The first concern is that while the testing methodology generally appeared adequate with controls in place, proper neutralization study, and adequate population counts, the lab did not abide by nor list standard testing references. Testing references as required by USP ensure that the methodology used by one laboratory can be understood and replicated by a second laboratory doing identical testing. Proper commercial microbiology analysis and reporting require references, and Robison's report lacks standard testing references.
Secondly, testing of the kill rate (efficacy) of the silver product was based on a single time point of "2 minutes." Reputable testing of kill rate would require at least 3 or more different time points for analysis. In other words, what is the kill rate at 1 minute, 2 minutes, 3 minutes, and 5 minutes? Is there an appropriate trend with a greater kill rate at 3 minutes compared with 1 minute? Without testing at multiple time points, there would be no conclusive evidence that the product had an effective kill activity that would increase over time. Robison's study based only at a 2-minute measurement lacks the multi-time data needed to confirm product efficacy.
Argentyn 23 has had a compendial study performed by an accredited third-party laboratory following referenced methods that demonstrated at 2 minutes a kill rate for Argentyn 23 of 99.996% with a log reduction of 4.5; this is clearly completely in disagreement with Robison's testing of Argentyn showing a kill rate of 47.1% with a log reduction of 0.28. That there can be such a great disparity between the results questions the validity of Robison's testing of the products. Without such validity and for the other aforementioned concerns, the article is called into question and is hereby retracted.
Jonathan Collin, MD