The following question was posed to several blood bank experts on the CBBS panel and also posted for comment by John Judd and John Case.

Blood bank laboratory reagents are to be quality-controlled each day of use to document acceptable reactivity. Do the panelists apply this principle to the QC of anti-C3b,C3d antiglobulin reagents? If yes, by what methodology? If no, why not?

The following ideas were submitted in response to the posed question. These opinions of the CBBS panelists are paraphrased and are being presented without attribution. They do not represent an official opinion or position of the CBBS. Also, the comments of John Judd and John Case (as expressed on the AABB-Online Bulletin Board) are presented at the end of this email.

1. One of the panelists and the panelist's Reference Lab Supervisor thought that it would be excessive to perform QC for anti-C3b,C3d reagent. That panelist said that in order to QC anti-C3b,C3d, one would need to buy C3-coated cells, and then asked "how do you QC these cells independently of the anti-C3"? The panelist further commented that (in the panelist's opinion) the only use for anti-C3 reagent was is in a DAT (Web Master Note; I do not agree with this statement as the reagent is also used sometimes in indirect AHG testing). Thus, according to the panelist, QC of this reagent was unnecessary, since "The only need for  anti-C3 is in DAT, which is helpful but not necessary for clinicians to diagnose hemolytic processes", and "QC does not replace clinical judgment."

2. The remaining panelists argued in favor of performing QC on anti-C3b,C3d reagent. One panelist commented that his laboratory purchased complement-coated control cells from Gamma, and that his laboratory used these cells as a control of for anti-C3d antiglobulin reagent.

3. Another panelist commented that their blood center reference laboratory performs quality control on anti-C3b,C3d reagent with complement Coombs Control Cells (C3b) that are purchased from Gamma Biologicals. The complement coated cells are expected to be agglutinated (1-2+) by antiglobulin sera that contain anti-complement; including polyspecific and anti-C3b/C3d reagents. The panelist went on to say that they check the anti-complement-containing reagents with these coated cells daily, and they also add coated cells to all anti-C3b/C3d tests that appear negative--like the Coombs control cells that are added to apparently negative antiglobulin tests to confirm that the anti-IgG is "active". The panelist concluded by stating that readers will find complete directions for performing this testing in the package insert. 

4. Finally, a panelist admitted that his lab had not been doing QC of the anti-C3 component. The query to the Editor made his lab realize that they should be doing this QC. The panelist discovered that complement check cells are available from Gamma Biologics and are not very expensive. The panelist agreed that it should be a simple matter to follow the manufacturer's instructions which should be similar to methods for IgG coated cells.

5. As stated above, the question was also posted on the AABB On-Line Internet Bulletin Board in order to solicit the input of John Judd and John Case, both of whom responded. I have exercised Editor's perogative, and edited out some of the quips that occur between these two individuals. 

John Case wrote that in the case of polyspecific Anti-Human Globulin, it is generally argued that to demonstrate continuing anti-IgG reactivity is to cover the anti-C3d component. If the anti-IgG component is working, then the anti-C3 must be working also. This argument has many convinced adherents, but is not very rational. The anti-C3 component has a significantly lower titer than the anti-IgG; hence, you could have your reagent so diluted (through leaving too much saline on the cells after the last wash), or so impaired by partial neutralization, as to diminish the anti-C3 reactivity while leaving the anti-IgG ostensibly intact. However valid that reasoning may be for polyspecific AHG, it does not legitimately apply to anti-C3b,-C3d, which was the subject of the question. One feels that one undoubtedly ought to be doing regular quality control, if for no other reason than that this is supposedly required by Standards. Speaking personally, I'd say it is especially important in the case of this specificity, because anti-complement tests are not, generally speaking, carried out well. Agglutination due to anti-C3b and anti-C3d develops slowly; yet there is still an expectation that it ought to come up at immediate spin. Even when the manufacturers' recommendation to wait for five minutes between adding the reagent to the washed cells and centrifugation is faithfully carried out, the resulting agglutinates are fragile, and are easily dispersed by the lack of gentle shaking. A positive control is a useful means of reminding the user what to expect. This reason could apply equally to Polyspecific AHG, needless to say. Where to get the necessary control cells? Well, you can make your own. There are methods in the literature, but they are admittedly rather tedious, and applying them may entail complications arising from the observance of good manufacturing practices, which most laboratories would (understandably) prefer to avoid. It would be hard to justify the hassle, particularly as your home-made control cells would have a very short storage life (due to being made without aseptic precautions). The alternative is to buy them.

To which John Judd commented that just because Standards says we should or must, does not mean it is relevant to patient care! If we need to perform daily QC on our anti-C3 reagents, then surely - by the same token - we need to perform daily QC studies on each antigen claimed to be present on reagent cells used for antibody detection and identification. How do we know that we haven't denatured S antigen on these cells by too liberal use of NaClO? We also need - by the same token - to show that our acid elution kits are still active (eg, we haven't inadvertently contaminated the product with NaOH). When can we stop this nonsense?

We rarely use anti-C3 reagents alone in antibody detection/identification tests. When we do, it is purely for academic reasons. We do use anti-C3 in the workup of a patient with autoantibodies or when immune-mediated hemolysis is suspected. The test with "monospecific" anti-C3 is invariably done in parallel with or after a positive test with polyspecific AHG. In our institution we use reagents from two different manufacturers for this testing. I doubt if there will ever be a situation in which 2 different polyspecific AHGs and two different anti-C3 will be inactive, for whatever reason, unless the whole rack of AHG reagents got fried or was left at RT for weeks on end (which it never is). Regardless of laboratory findings, the diagnosis of immune hemolysis is based on clinical observations. I say that this sort of QC testing - although claimed to be essential by those who have nothing better to do but conjure "quality processes" - is sheer and utter twaddle!

To which John Case commented that in his view no daily quality control should be done that does not provide truly meaningful information. Applying that principle unquestionably precludes testing the integrity of antigens on Reagent Red Blood Cells, which offers nothing useful besides confirming the reactivity of that antigen with that example of its corresponding antibody in the hands of that particular bench worker.

One could argue, however, that to run control tests on day of use with Anti-C3 reagents does at least (as I argued) give the user a hint of what is to be expected in terms of a positive result. Beyond that, I don't suppose it tells you much. For instance, "day of use" testing doesn't allow for differences of manipulative skill in resuspending the cells, which is of particular relevance when agglutination is as fragile as that produced by anti-complement components. I am at something of a loss to know why testing any sample needing to be tested for a coating of complement with two examples of polyspecific and two examples of anti-C3 would be easier, cheaper and more informative than confirming the reactivity of a licensed anti-C3 with a positive control.

To which Mr. Judd admitted that Mr. Case was absolutely right that our duplicate testing seems excessive, but we have had trouble with these products in the past, ever since one manufacturer ceased marketing blood bank reagents. We have seen discrepancies between manufacturers, that are not always consistent, and I don't quite know what to make of it. In any event, when I have nothing better to do I need to review our most recent data. As to what we do when the test with one reagent is positive and the other negative, we report the results as equivocal. I have never been convinced that such discrepancies are due to reagent inactivation or deterioration, the reagent that is non-reactive with the sample in question still reacts with other complement-coated RBCs. We would, of course, review the clinical findings on the patient and discuss our results with the attending physician.

Editor's NOTE:  It seems that Mr. Judd's comments support the importance of QC-ing anti-C3b,C3d reagents (even if Mr. Judd doesn't realize it)!!!!

ADDENDA March 8, 1999

The following is additional information on the QC for using anti-C3 AHG reagents.  I wish to thank John Case for the effort he has expended to provide this information.  What follows is extracted from (and ever slightly edited) from the AABB On-Line Special Interest Group. 

Preparing C3-coated cells: Tricks of the trade (and threaded discussions)
by John Case, FIBMS, 2/9/99

Ask anyone with experience in coating red blood cells with complement, and you'll be cautioned that the procedure is not nearly as simple as it seems, even when written directions are followed to the letter. Cells coated with C3b are prepared by the alternative pathway using freshly drawn blood. The cells so coated are then treated with trypsin to simulate enzyme action in the circulation that cleaves C3c and leaves the cells coated with C3d.  Subtle variables in executing the procedure can influence the performance of the final product, which may be reason enough for users to observe differences in the strength of agglutination between separate manufacturers. Even more significant, perhaps, is the serum used as the source of complement, which is a variable that can introduce lot-to-lot differences among products of the same manufacturer. Then, last but not least, is the human element, which may be the most significant source of variability.

A few years ago, John Case served as co-convenor of a working party appointed by an Expert Panel of the International Committee for Standardization in Hematology (ICSH) and the International Society of Blood Transfusion (ISBT) to select a suitable polyspecific AHG reagent to serve as a new international standard. The task was accomplished by circulating seven reagents containing a blend of polyclonal anti-IgG (rabbit) and monoclonal anti-C3 (murine) to twelve participating laboratories for evaluation by methods laid down in the accompanying instructions. In comparing anti-C3 reactivity, participants were instructed to use indicator cells prepared both by the conventional and the alternative pathway, and to apply them both as EC3b and as EC3d (prepared by trypsin treatment). Everything was supplied. The buffers, the trypsin, the monoclonal (IgM) anti-Jka kindly donated by Bioscot Ltd (to be diluted beyond its titration endpoint as a direct agglutinin) by which to trigger the conventional pathway, and detailed directions. The complement-coated cells were to be tested with serial dilutions of the seven antiglobulin reagents, both with and without delay before centrifugation.

The results from the eleven (of twelve) participants that undertook anti-complement testing showed substantial discrepancies with C3-coated cells. A few participants encountered difficulty in observing agglutination of their C3d-coated cells at all in an immediate spin test, regardless of the means used to coat the cells. A five minute delay improved these results, but endpoints reported by these participants were invariably lower than those reported by those that obtained endpoints at immediate spin. Titers with C3b-coated cells were rather better overall, but there were still startling discrepancies between individual participants. Even the one reagent that stood head and shoulders above the other six in regard to its anti-C3 activity (which was subsequently rejected as the international standard because the Expert Panel adjudged it to have too potent an anti-C3 component to be realistically attainable by manufacturers whose products would have to be compared with the standard) performed poorly in the hands of some participants.

It is true that subtle differences in the procedure of making the coated cells in the participating laboratories may have been a factor in producing variable strength of agglutination during the study. The conventional pathway with anti-Jka is significantly less prone to variability than the alternative pathway, however, as long as everybody was using the same antibody at the same dilution, which supposedly they all were.

The foregoing remarks illustrate how the manner in which the deposited cell button is resuspended can influence the result, how C3d-coated cells can be expected to react less strongly (and exhibit a lower titration endpoint) than C3b-coated cells, and how a short delay before centrifugation allows the reaction to reach some kind of stability and, thereby, enhances the reaction.

Now for my opinion about the value of QC'ing anti-C3 reagents, whatever it is worth. If we are going to attach significance to the results we obtain in testing with Anti-C3 (and one has to presume we are if we are going to the bother of doing the testing at all), then we do need to undertake quality control as a means to confirm that the reagent is working properly. There's more! Testing "on each day of use" does not go far enough. Based on the inescapable truth that the results of this particular test vary so considerably in the hands of different workers, a parallel positive control test is needed with every batch of tests. This is needed to provide evidence that the anti-C3 component is working in the hands of the worker undertaking the particular batch of tests, and also illustrates the kind of result that is to be expected. Finally, testing with Gamma's C3b-coated cells is not a satisfactory test of anti-C3d activity. This product was designed for a Coombs Control test, similar to the confirmation of negative tests with IgG-coated cells. Whether or not that is a worthy purpose may be debatable, but views on that score are not germane to our present discussion. Anti-C3d does indeed react with C3b-coated cells, since C3d is a component of C3b. The only exception would be a monoclonal anti-C3d directed at an epitope that becomes available only when C3c is cleaved. But an "anti-C3" reagent containing potent anti-C3b but no anti-C3d component at all will give a reassuring positive result with C3b-coated cells.

Quite a lot of quality control testing is mindless, useless, and serves no valid purpose, Testing reagent red blood cells daily for some arbitrarily selected antigen stands out as a case in point. As I keep saying, a justification for using a positive control when testing with Anti-C3d is that (quite apart from providing evidence that the reagent is working) it gives the user an indication as what a positive reaction ought to look like. Most anti-complement testing is not very reliably done. People shake too hard, or fail to allow the reaction to reach stability by postponing centrifugation in accordance with manufacturer's instructions.

While I am inclined to agree that the diagnostic value of demonstrating C3d on the cells of a patient may be dubious, the testing concerned is being done, and it is often done badly. I can see no legitimate parallel with ABO and Rh reagents, which usually yield strong agglutinates that even the most vigorous shaking does not disperse. In such situations, day-of-use testing is perfectly fine. An exception, perhaps, is testing with Anti-C, which ought to be done with cells representing C as a product of the Rz or the ry gene, or (at a push) the r' s gene, as being likely to yield weak agglutination. Here, as in all situations where the reaction may be relatively weak (and thereby susceptible to significant variation in the hands of individual bench workers), it would improve the quality of patient care were a positive control to be run with each batch of tests, rather than simply on each day of use. Testing with Anti-e is a perennial problem. It always has been. There are so many variant forms, some of them definitely qualitative, that truly meaningful quality control is hardly possible. As for antibody screens, the only meaningful quality control is to check for hemolysis and discoloration. My view on this is well known. Choosing one antigen to test is a wanton waste of time, effort and materials.

Please submit comments to the e-Network Forum.

Ira A. Shulman, MD
CBBS e-Network Forum Editor & Moderator