Implementation of a terminal analytical control in cytotoxic reconstitution unit (CRU)

Introduction
Following the increased activity of our production of chemotherapies, we aimed to enhance the safety and our quality control requirements. Among the existing choices such as dosage, video or weighing, we opted to introduce a terminal analytical control, in addition to the existing visual double-check by pharmacists.

Objective
Validate the routine use of the qualitative and quantitative control by UV-visible spectrometry DrugLog^® performed by Pharmacolog^® group for the release of preparations in our CRU.

Materials & Methods
The device calibration was performed by molecule/solvent groups without discrimination of monoclonal antibodies. In fact, 10 or 20 samples were required by Pharmacolog^® experts to validate our calibration curves. An analysis of the first results was conducted from November 2022 to May 2023 via the DrugLog^® internet platform with extraction of non-conformities (NC). The device was evaluated by the pharmacists and pharmaceutical assistants through a satisfaction questionnaire.

Results
Fourteen molecules, including 4 antibodies, have been validated in routine. Analysis of the results reveals an absolute NC rate of 5% for 869 samples collected. Two types of NCs are observed: either the molecule or concentration is not identified (No match), or the concentration is outside the acceptance threshold set at +/- 15% (Fail). The average score for the control device is 15.2/20 (n = 15; min-max = 9-17/20).

Discussion
The reported NC rate remains higher than the rate of 1.3% found in the study carried out by the University of Geneva, which suggests areas for improvement. However, this rate drops to 1.5% if explainable NCs are extracted (product selection error, incomplete cuvette, incorrect/omitted homogenization; n = 33). The unexplainable NCs (n = 13) led to the systematic preparation remanufacturing. This survey has allowed us to discuss avenues for improvement with Pharmacolog^® and in particular to import the absorbance curves of our samples on the internet platform and to provide assistance in the interpreting of artefacts such as microbubbles. Among the reasons for dissatisfaction mentioned, the difficulty of distinguishing validated molecules from those being calibrated complicates the interpretation of results in routine. The addition of a color code has therefore been suggested to the Pharmacolog^® group. Moreover, a guide detailing the explainable NCs has been made available to the pharmacist with the necessary steps to take before concluding a true NC.

Conclusion
To conclude, this quality control device is an asset in securing the circuit of injectable chemotherapies prepared by our CRU. The responsiveness of the Pharmacolog^® group to our various requests and the time required for analysis are favorable for its routine use. However, the sustainable financing of this device remains a challenge.

See the poster