Abstract
(Englisch)
|
General goal
It is aimed at replacing the rabbit pyrogen test by one or several in vitro methods. To achieve this goal, several laboratories have optimized their method of alternative pyrogen testing,
and an inter-lab transfer has been made. Thus, each of the 6 methods is performed in the developing lab (leading lab), and in two receiving labs, one being regarded as the expert lab, one as the na'ive lab, with regard to the transfer of this method.
Progess
In a first phase, each of the developing labs has optimized their method and has
summarized the procedure in a standard operating procedure (SOP). In a second phase, a transfer of the methods between the labs has been made. Thus, three labs now have the method developed in Berne implemented in their environment and have been familiarized with the method. As problems of transfer occurred, the SOP's have been refined. In a third phase, agents regarded to be clean were distributed such that each of the 6 labs worked
with the same material, and data were compared. In a fourth phase, the robustness of the methods were evaluated by a variety of criteria being defined by the statistian accompanying the project.
The fifth phase was the prevalidation preceding the final validation phase. In the prevalidation, only E. coli type endotoxin was used as a test substance. It was tested for pyrogenicity when dissolved in a variety of test solutions. As test solutions might interfere with a given test system, it was first tested by the developing labs which is the highest concentration of a set of test solutions at which a recovery of a signal derived from E. coli type LPS was obtained within predetermined borders (50% to 200%). This recommended test dilution has to be multiplied by the dilution of the test material in the assay, which differs for the 6 assays concerned. This yields a final dilution at which the test solution can be reliably tested for recovery of a given concentration of LPS (termed spike). After the recommended test solution was communicated, 4 unknown spike concentrations had to be admixed to the test dilutions in a blinded fashion. The spikes were calculated such that their final dilution corresponded to a threshold of 0.5 EU/ml and 1 EU/ml final concentration,
which was regarded as a borderline concentration to which the most sensitive rabbits might react. The experimenter did not know whether the spike was contaminated or not. The data were evaluated with regard to sensitivity (how many false-positives) and specificity (proportion of false-negatives). Data evaluation showed that overall, the sensitivity was good (better than for the rabbit test) The results showed a good overall sensitivity and an excellent
specificity. Each of the 6 methods performed well and was regarded as suitable to replace the rabbit test.
Prooress of THP-1 model developed in Berne
The THP-1 model developed in Berne has experienced an evolution over time, and looks now quite different than at the beginning. We are using another clone (THP-1 2A9) , and another way of evaluation. Thus, we determine TNF by a one-plate ELISA, i.e. the plates in which the THP-1 cells are cultured when exposed to pyrogen test substances are already coated with anti- TNF-antibody. In a 2nd step, cells are removed by washing, followed by the addition of detecting antibodies. Current data suggest that the limit of detection of the one- plate assay is lower than with the two-plate assay previously used. One problem to be solved is the choice of the substrate that needs to be changed since the presently used substrate does not allow the 0.0. readings to be standardized.
Overall, the THP-1 clone method developed in Berne not only was capable of detecting pyrogens accurately. It proved to be a viable alternative to assays in which whole blood or blood-derived cells are used, since not all laboratories have easy access to whole blood. To comply with the original time plan, the personnel participating has to be fortified, however.
Proaress made reoardino whole blood assay (format Constance)
Berne was assigned as the expert 'receiving' lab of the whole-blood test system developed in Constance. This test was implemented in Berne in spring 2001. There was some initial controversy as to whether in this method, preselected donors or random donors should be used. The assays carried out in Berne using this test system were largely successful.
Future directions
Presently, it is discussed whether a generally applicable prediction model should be used, and if yes, which one. While the actual logistics and the selected agents in the actual validation study are being worked out by the task force, the general outline has been agreed on. There will be
(A) a fixed set of substances set up by the task force and blinded,
(B) a number of 'difficult' non-endotoxin pyrogens from which we wish to know whether the different assay systems are capable of picking them up under blinded conditions,
(C) A number of potential pyrogens from which we want to know whether the different assay systems recognize them or not.
Before part (A) can be set up, the level of interference for the test systems must be known, something to be evaluated by the developing labs.
Dissemination of the data
It is planned to prepare a publication jointly, by the participating 6 labs and ECVAM. In addition, two further publications will be submitted shortly. In one, THP-1 cells are compared with human macrophages and dendritic cells with regard to the spectrum of pathogen- derived molecules they react to. In a second, the newly developed one-plate assay will be compared with the established two plate assay, and the new clone presently used, will be described.
|
