Tips and Tricks for Validation of Quality Control Analytical Methods in Good Manufacturing Practice Mesenchymal Stromal Cell Production
Table 1
Validation strategy for “safety” detection methods.
Test method
Ref. Ph.Eur./ICH
Validation steps
Evaluated parameters
Acceptance criteria
Microbiological examination
Ph. Eur. 2.6.27/ICHQ2
Analysis on CBMSC () and BMMSC () spiked with 1–10 CFU and 1–100 CFU of each microorganisms Volume of inoculum: 1% of the total batch volume Cryopreserved (CBMSC) and fresh (BMMSC) sample validation
Accuracy
Microorganism growth of medium alone comparable in the presence of the product as confirmation of antibacterial activity of the product
Specificity
Detection limit
No growth in the negative controls (specificity)
Limit of detection: 1–10 CFU
Precision
Same results for each validation run performed by at least two different operators
Bacterial endotoxin test
Ph. Eur. 2.6.14 (method D)/ICHQ2
Assurance of standard curve criteria
Linearity
of standard curve ≥ 0.980
Study of the product (according to clinical use)
EL and MVD calculation
NO-interfering dilution < MVD
Test of interfering factors
Accuracy (%)
50% < spike recovery < 200%
Specificity
Onset time of negative control and sample(s) no spike > onset time
Precision
CV (intra-assay) ≤ 10%
Sensitivity
× chosen dilution
Adventitious viruses analysis
ICHQ3/ICHQ5A(R1)
Preliminary phase
Cell number, extraction, and amplification conditions setup
No inhibition on extraction and amplification for respiratory, CMV, and EBV viruses, with the kit usually used for standard biological diagnostic samples
MSC spiked with two viral loads of adenovirus, enterovirus, CMV, and EBV
Specificity
Detection unequivocally of the specific viruses in BMMSC and CBMSC which may be expected to be present
MSC () spiked with a viral load ten times the cutoff of sensitivity declared by the manufacturer
Accuracy
For CMV/EBV (quantity analysis): quantity of target DNA present in the sample reaction (gEqu/reaction) in spiked MSC samples similar to positive control (−20 ≤ accuracy error ≤ +20)
Detection limit
Ability to detect the viruses in spiked MSC near the cutoff of the kit
