A question was posted to an online Q&A board directed to CAP in December 2012. The question read, in part, "I understand that the CAP Laboratory Accreditation Program recently announced that negative controls are no longer required for immunohistochemistry ... "
It was something of a leap of misunderstanding from what was actually stated in CAP's July 2012 rewording of its anatomic pathology checklist on the use of negative controls. Certainly CAP relaxed the imperative for negative reagent controls (NRCs) by stating, "Immunohistochemical tests using polymer-based detection systems are sufficiently free of background reactivity to obviate the need for negative reagent control and such controls may be omitted at the discretion of the laboratory director." So the answer, provided by Regan Fulton, MD, a member of CAP's Immunohistochemistry Committee, to the aforementioned query offered detailed clarification: Both positive and negative tissue controls are still required by CAP and CLIA to assure stains are working properly. However, as concerns negative reagent controls, when newer "multimer"- and "polymer"-based detections are used, a negative reagent "is not generally helpful."
Fulton went on to explain, "The negative reagent control is a replicate patient tissue slide in which the primary antibody is replaced with either another antibody of the same species and isotype or, in some laboratories, simply diluent. This slide is generally treated with the most aggressive antigen-retrieval protocol used among the various tests in a given case. This control is designed to indicate the presence of non-specific binding of detection reagents due to the presence of endogenous biotinae as may be seen in a variety of normal and neoplastic tissues, most notably liver and kidney."
The answer continued to state that because newer multimer-based and polymer-based detection reagents are sufficiently free of background reactivity, a negative reagent control is not necessarily helpful. Fulton said, "The CAP Immunohistochemistry Committee has concluded that the value added by routine use of a negative reagent control does not outweigh cost in terms of labor, reagents, materials, space on staining instruments, and limited diagnostic tissue . However a common-sense approach is required. Negative controls are indicated in cases where there is unexpected staining and should be added . at the discretion of the pathologist."
However, in a 2013 Dark Daily white paper titled "Identifying False-Positive Results in Immunohistochemical Assays: The value of negative reagent controls in the clinical laboratory," authors Eric Walk, MD, FCAP, and Peter Banks, MD, FCAP, both associated with Ventana Medical Systems, take CAP's position somewhat to task. They raise a provocative point: "Economic concerns, especially with a constant push for greater efficiency and cost-savings, have made the elimination of NCRs an attractive option for decreasing costs and laboratory time. However, there are significant risks to the elimination of NRCs, primarily being compromised patient care."
They note that the "purpose of NRCs is to avoid the misinterpretation of false-positive staining as 'true' positive staining. Notably, false-positive staining can result from any one of several diverse pre-analytical and analytical factors, including the properties of the specimen itself, specimen processing methods and specimen storage conditions. Many of these factors are both unstandardized and beyond the control of the clinical laboratory."
The authors also point out that when "very rare foci of staining are considered diagnostic for positive interpretation, or when cancer patient eligibility for a proposed treatment depends on a particular IHC result" greater caution is needed beyond that which is needed for commonly encountered diagnostic applications. This is particularly true in the use on companion diagnostic assays, they note, begause the decision to treat is based on the IHC diagnostic result.
Because of these (and other) critical considerations noted in the paper, the authors said that Ventana Medical Systems will continue to recommend the use of NRCs as a "necessary component of proper interpretation of IHC assays."
An Opinion from the Field
Kathryn McIver, MT, chair for the American Society for Quality in Denver, Colo., formerly was quality coordinator for MetroPath where she was also the lab's Lean Six Sigma project leader. She champions efforts to streamline processes, and commends CAP for being "leaders in advancing technology to ensure compliance with practices that provide the best level of care for patients." However, she, too, has some uneasiness about the relaxation of NRCs.
"The change in the use of reagent controls really is moving us from having a control on every single test to a spot check. It's a little scary to me," she told ADVANCE. "As a quality coordinator, I would not favor changing the way we had been doing things. Yes, there has been quite a bit of advancement in technology where we have more reliable testing, and we are certainly going toward automation - which will help standardize the way things are done. However I personally would not be comfortable with going to [randomized controls] without quite a bit of structure in my own laboratory to make sure that I would be confident in the validity of testing. I would want to do our own double-blind internal validation for any testing for a high-risk patient, because when you are doing oncology testing there are so many external factors that could impact staining - specimen prep, specimen handling, etc."
McIver admitted that there is a compelling financial argument for cost savings via a reduction in the number of NRCs conducted by a lab. But she also pointed out that these savings would be offset by initial investments prior to the reduction of the NRCs.
"For many labs there is necessarily going to be a large initial investment to switch over to a system where you could be confident in results when not running the NRCs on every single specimen. Labs may require greater instrumentation to take advantage of advancements in technology, tests will need to be more specific, so there could be a higher cost per test; and certainly staffing must be brought to a very high level of proficiency. If you are an oncology lab and this is what you do all day, this might well be a cost saving in the long run. But if you are a generalist laboratory or a hospital laboratory, this may not be a cost-saving activity."
McIver, insistent that "patient care is ultimately our major concern," added that labs already understand the diagnostic value of NRCs, but do they really know the effect of not using NRCs each and every time? "We really don't have a good understanding of how that is going to change the interpretation of results. I can't speak for every quality coordinator out there, but from my standpoint I would much rather spend the extra money than ever take the chance of having a Type I error. I would always want to do a very thorough test."
As compelling as McIver's arguments are, there is still another perspective worthy of consideration.
Ralf Huss, MD, PhD, chief medical officer at Definiens and co-founder of the biotech company APCETH, has more than 20 years of training and experience in histopathology and cancer research. Huss told ADVANCE that he generally agrees with CAP's change of language. While he, too, is adamant that patient safety is the most important issue at stake, he is not convinced that status-quo use of NRCs always "deliver."
"If you try to have a negative control, you have to be under the assumption that the reagent that you are using is working, otherwise you would not get a true negative control," Huss explained. " If you get some tissue that has been underfixed or overfixed, or has been put near an oven, or into the bright light, or any number of other weird conditions, you might get a negative result. And if you don't prepare a specimen properly, then a negative control simply doesn't help. It is so easy to [inadvertently] make any kind of tissue 'negative.' The fact of the matter is, there are so many causes for getting false negatives that actually a negative control really doesn't help."
This variable handling of specimens, combined with what some might consider the "interpretive art" of pathology leads to a great deal of subjectivity, said Huss. "Therefore, the NRCs might give you a false sense of security in the 'validity' of incorrect results."
Huss believes that CAP is on the right track, encouraging labs to use standardized, well-established assays on certified platforms and run according to the manufacturer's instructions,". instead of trying to do something on your own - some kind of negative testing that gives you a false sense of security which does not exist. I am a pathologist myself, so I feel I can say this: We are still years behind, for example, clinical chemistry. They have standardized everything. Everything is on a platform. They run controls. They are highly sophisticated. But most of the pathologists - rightly so, or not - consider pathology a type of art. To secure patients' safety, we must come to a certain degree of perfected organization, of standardization, quantification and objectivity. There is still a lot of art to it - developed from knowledge which is ingrained in the experience of the pathologist. But we also have to apply a more standardized approach and not leave so much room for subjective assessment by pathologists."
Huss believes that previous protocols were inherent with the possibility for error. "But CAP is trying to carefully move us to the right direction . A lot of lab directors and pathologists still believe this kind of negative control will still work in their hands. But CAP's influence is nudging them into a carefully thought-out movement in the right direction."
Valerie Neff-Newitt is on staff at ADVANCE. She can be reached at firstname.lastname@example.org.