exlimey

Bottom line: If a facility "manufactures" a reagent, it is obliged to prove that it functions as designed and can be made repeatedly/reliably. However, the rules and regulations that apply to licensed and registered reagents do not necessarily apply to "home-brew" materials. Providing the material will only be used in-house and not distributed, there is no need for "parallel testing" of any other formulation, no matter how crude the manufacturing process appears to be. Such testing is only required for licensed or registered reagents and devices (under the umbrella of Design Control). I'm assuming (correct me if I'm wrong) that the PBS used in preparation of the DTT is made immediately before use, and therefore the stability of it is irrelevant. Stability of the finished reagent may be important if the "manufacturers" wish to assign an expiration date. This may not be necessary, since in the case of DTT, controls are included EVERY time it is used - the operator has a current and immediate indication of whether the reagent worked or not. Validation is essential to provide confidence when it is impossible to directly prove that a process has worked. Think of vaccine manufacture: One cannot test every vial/dose to see if it meets specifications - all of the product would be compromised. When integral testing proves that a process works each time, validation requirements are minimized.

What do you mean by "extensive in-house validation" ? What is "in-use testing" ?

In my opinion: Yes, you can. That statement on chemicals is meant to tell you that you can't use it in a medical or nutritional fashion. You will not find a package insert for any raw chemical - the supplier has absolutely no idea what the buyers are going to do with the materials. If you buy plain old sodium chloride (NaCl), it doesn't have a package insert, exactly for the reasons stated above. Anything anyone is doing with DTT and/or other exotic chemicals in the Blood Bank realm is completely out of the typical regulated environment. These chemicals assist in a complex investigation, they are not making a diagnosis. Sometimes it is necessary to go beyond the use of licensed, registered, validated reagents to best serve a patient. That being said, the DARA issue has brought DTT use into some routine labs. Complex serological investigations should be left to experts - the high level Reference Laboratories, who understand the pros, cons and limitations of the specialized reagents they use. End of rant.☺

You may be referring to trypsin-treatment of the red cells (screening cells). Apparently CD38 is destroyed/inactivated (along with Lutheran system determinants) but Kell system antigens remain intact. Other blood group antigens are also affected by trypsin, so I think the modified approach involves testing the patients' samples against both DTT-treated and trypsin-treated cells. To further complicate matters.....manufacturing a reliable, consistent trypsin reagent is VERY difficult. The enzyme activity of source material varies immensely and, as with other enzymes, stability is a problem.

You performed stability testing on your home-made material ?

Feisty rainbows? Some grayling? I'm very jealous - haven't made it to Alaska yet.......

I concur. A good writer should be able to finagle that logic into a validation plan. Getting hold of said cells may still be problematic, but at least it might be simpler than an expedition to catch the Loch Ness Monster.

First, a disclaimer: I am not a Regulatory expert. Perhaps you're over-stretching? I don't think you're required to validate the new system in the sense that the manufacturer has to do for licensing - that's a LOT of work and an attempt to cover every variable imaginable. In your case, perhaps just proving the system works in your facility is enough. That would probably only need to include a couple of examples of weak-D, rather than the whole gamut.

Well said. The world is so "dangerous" these days, it's a wonder we're still around.

Contrary to the manufacturer's instructions ? That's really living on the edge. ☺

I must confess.....I was poking the bear. I have worked with LN2 for 25+ years and personally believe the risk of asphyxiation is highly over exaggerated. Any room with adequate ventilation (and that's a very loose term) is safe. Most interactions with an LN2 environment are short term, except perhaps for the worker(s) at the NHSBT National Frozen Blood Bank where special ventilation systems should have been designed into their operation. I was expecting answers involving cryogenic burns, too.

What, precisely, are the "H&S" risks?

Anything is possible, but the screening cells are designed with the intent to detect all (common) antibodies, including the antibodies traditionally considered "colds" - MN, Lea/b, P1, etc. It is remotely possible that the frequencies/incidences of an antigen have come together perfectly to yield these results, but you'd have to be very lucky or unlucky, depending upon your point of view.

If the screen is negative by IS, it rules out the presence of a generic "cold". Are you only seeing this phenomenon during an IS crossmatch ? Are you only crossmatching group A units ? Does the same "incompatibility" happen if you crossmatch group O units ? Typically antibodies detected prior to an antiglobulin phase are IgM. These do not necessarily "carry through" to the antiglobulin phase, especially if they have limited thermal amplitude.

I don't know why the Control is positive, but the actual test is negative. In theory, the anti-D and control should be a matched set - they should be formulated exactly the same, except for the presence of the anti-D. Every manufacturer has "secret ingredients" - potentiators and other chemicals that stabilize and enhance the antibody reactivity. The controls, if supplied, should include the same ingredients. If you are using a control from a different manufacturer or product line, it might explain your findings. Are you saying that a monospecific anti-complement reagent is reactive with the patient's cells? It's not unheard of for warm-autos to bind complement, but since this looks like an autoanti-e, it would be highly unusual.