By Agnes Shanley, Editor in Chief
As the concept of real-time release becomes more important, the harmonization of on-line total organic carbon (TOC) analytical methods is becoming a contentious issue.
Recently, GE Analytical Instruments proposed a new standard practice to ASTM’s E-55 PAT committee, members of which are now commenting on the draft. The stated goal of the proposal is to develop a science-based method for validating on-line total organic carbon methods for pharmaceutical-grade water. Its wording emphasizes the potential for problems when moving from lab-based to on-line systems, and the lack of common science-based approaches to method validation. It also cites ICH Q2B, which set limits of quantification and detection for TOC systems without dictating testing methods.
So why has it enraged so many experts in the water analytics field?
TOC testing is a delicate thing, and results can vary tremendously, even for the same installation, depending on trace amounts of compounds in the source water and on such factors as diffusion effects and the catalytic reactions that can occur in uv-excited water, according to Todd Blonshine, senior scientist at Hach Ultra Analytics (Cary, N.C.).
In a subject area where the world’s top technical experts are all aligned with vendors, and three vendors, GE, Hach Ultra, and Mettler-Toledo Thornton (Bedford, Mass., the latest entrant into the market) compete for shares of the on-line TOC business (Japan-based Shimadzu has introduced an online technology as well), the controversy promises to continue. In the end, it is up to industry professionals to weigh the facts and decide for themselves.
Broadly, there are two basic methods used for on-line TOC analysis:
- Continuous and direct conductometric (also known as Direct Conductometric (DC) with stop flow or continuous flow oxidation technology);
- Membrane conductometric (MC) system with UV alone, or with UV and persulfate oxidation.
Each has its own strengths and weaknesses, which will vary depending on installation, and which the astute user can adjust for. DC methods are fast, and facilitate connection to process control systems. They are sometimes prone to high readings in water that contains high concentrations of chlorine or some organics, but this problem can be addressed through pretreatment.
Membrane methods do not show false positive responses in water with high chlorine concentrations, but require low ambient pH level to work effectively, and take readings at a slower rate.
Currently, for much of the world, the standard for both lab and on-line TOC instrumentation, including systems used for real-time release of pharmaceutical-grade waters, is USP 643. Developed in the early 1990s after five years of discussions involving Sievers (now GE) and Anatel (now part of Hach), as well as extensive tests on both on-line and laboratory systems, it became official as an optional test to oxidizable substances in November 1996 for purified water and WFI. In 1998, the oxidizable substances test was deleted and TOC became the only official test for organic impurities, so 643 became the definitive standard.
USP 643 establishes acceptance criteria for any on-line or laboratory TOC instrumentation for pharmaceutical-grade water, explains Anthony Bevilacqua, head of R&D for Mettler-Toledo Thornton, and chairman of USP’s Water Committee. These criteria are based on detection limits, the instrument’s ability to distinguish between organic and inorganic carbon, as well as a suitability test designed to challenge existing measurement technology, involving one easy to oxidize compound, sucrose, and one hard to oxidize compound, 1,4 benzoquinone.
Six years ago, the European Pharmacopeia harmonized its regulations to 643, and China is expected to adopt 643 soon, according to Hach Ultra’s Blonshine. Until very recently, Japan had only recognized its own standard, JP 15, (for more on JP 15, see www.pharmamanufacturing.com/articles/2006/236.html?page=3) which requires different suitability agents and uses combustion-type analyzers.
“The issue of online TOC measurement for pharma grade water is still controversial, and there is a long way to go before there is true international standardization,” says Kazushige Morimoto, Ph.D., Director of JP’s Division of Scientific Affairs.
However, JP made some revisions to JP 15 this year, which would take effect next year. Although the language of JP’s standard notes some potential issues with DC methods, it ends the paragraph by stating that, “as long as the water is sufficiently pure, meeting USP 643 or EP 2.2.44 will meet JP 15 requirements.” Morimoto explains that “sufficiently pure” means meeting water for injection (WFI) standards, which JP currently defines as 0.50 mg/L of total organic carbon.
This is the same level established within 643, so, some say, Japan has essentially harmonized to USP 643, although experts within JP are discussing reducing the WFI limit to 0.30 mg/L next year.
Experts within GE say that they have observed that false positives can be seen in DC readings under certain conditions. On p. 20, we present a paper written by Jon Kauffman, Ph.D., of Lancaster Laboratories, presenting these data. In light of these findings, GE has proposed that new test chemicals be added to the two that USP specifies, to address potential issues with DC methods. The company had proposed that nicotinamide be added as a difficult-to-oxidize reference, and JP had initially added that substance, but has since crossed it off the list.
Others believe that USP 643 compliance is sufficient. “USP’s 643 suitability testing already addresses the issue of individual system variability; there is no need for additional testing chemicals,” says independent Colorado-based consultant, Nissan Cohen.
Specifying new test chemicals, they say, would require the entire industry to revalidate all its online water TOC systems, an expensive undertaking that, Hach’s Blonshine believes, would result in little improvement in either water or product quality.
There are also questions about whether an E-55 standard is needed, and whether E-55 is even the right forum. Cohen does not see the proposed method as a PAT issue. Instead, he says, the topic should be taken up by ASTM’s D19 committee, which focuses exclusively on water analysis methods. In fact, suggests Blonshine, the issue could effectively be addressed using FDA’s comparability protocol.
The world of online TOC area is a small one. Richard Godec, Product and Market Development Manager at GE, was a cofounder of DC vendor Anatel, now Hach Ultra. Consultant Cohen formerly worked for GE. In a technical area where vendors exert so much influence, it is essential that pharma professionals take a broad view of risks vs. benefits for any online TOC water system. They should base validation and other decisions on what they see or expect to see in their water supply, on available pretreatment technology, and on such factors as operating costs and speed of response.