A Quality Risk Management Model Approach for Cell Therapy Manufacturing

International regulatory authorities view risk management as an essential production need for the development of innovative, somatic cell‐based therapies in regenerative medicine. The available risk management guidelines, however, provide little guidance on specific risk analysis approaches and procedures applicable in clinical cell therapy manufacturing. This raises a number of problems. Cell manufacturing is a poorly automated process, prone to operator‐introduced variations, and affected by heterogeneity of the processed organs/tissues and lot‐dependent variability of reagent (e.g., collagenase) efficiency. In this study, the principal challenges faced in a cell‐based product manufacturing context (i.e., high dependence on human intervention and absence of reference standards for acceptable risk levels) are identified and addressed, and a risk management model approach applicable to manufacturing of cells for clinical use is described for the first time. The use of the heuristic and pseudo‐quantitative failure mode and effect analysis/failure mode and critical effect analysis risk analysis technique associated with direct estimation of severity, occurrence, and detection is, in this specific context, as effective as, but more efficient than, the analytic hierarchy process. Moreover, a severity/occurrence matrix and Pareto analysis can be successfully adopted to identify priority failure modes on which to act to mitigate risks. The application of this approach to clinical cell therapy manufacturing in regenerative medicine is also discussed.     

Editorial

Ohne Zusammenfassung     

Medical impairing injuries among Swedish firefighters

Using insurance data, medically impairing injuries sustained by 111 firefighters were analyzed. The injury incidence per year among full-time working firefighters was 39 per 10 000, and among part-time working firefighters 6 per 10 000. More than half (51%) were injured during fitness training, 14% during job training, and 13% during emergency response, while 4% were injured in traffic crashes and 18% in other injury events. Soccer and floorball caused 82% of all fitness injuries, of which more than half were meniscal or sprain injuries of knees or ankles. None of those injured during emergency response suffered from burn injuries, however during job training 4 sustained burns. Less injurious fitness training and practice activities could greatly reduce impairing injuries among firefighters and consequently reduce economical losses and disability.     

Bear ye one another’s genetic burdens: the price of diversity and complexity

Genetic variability and diversity are the result of a mutation-selection balance that acts permanently within and between species. The presence of deleterious mutations is a necessary consequence of this process and thus “the price paid by a species for its capacity for further evolution” (Haldane 1937, Am Nat 71:337–349). Recent estimations of mutation rate in the human lineage has revived the debate as to whether the high number of deleterious mutations poses a severe problem for the future of mankind. Theoretical considerations allow a scenario in which the survival of the human race is maintained by truncation selection of deleterious mutations that removes as many mutations as appear anew in every generation. In this case our genetic burden is carried by those individuals that suffer a genetic death resulting from random distribution of deleterious alleles. Nevertheless, one has to ask whether the mutation rate may set absolute limits on the complexity of a species.     

Modified Wald statistics for generalized linear models

Summary: Wald statistics in generalized linear models are asymptotically ² distributed. The asymptotic chi–squared law of the corresponding quadratic form shows disadvantages with respect to the approximation of the finite–sample distribution. It is shown by means of a comprehensive simulation study that improvements can be achieved by applying simple finite–sample size approximations to the distribution of the quadratic form in generalized linear models. These approximations are based on a ² distribution with an estimated degree of freedom that generalizes an approach by Patnaik and Pearson. Simulation studies confirm that nominal level is maintained with higher accuracy compared to the Wald statistics.     

Inferences concerning interpersonal choices given anonymously

It may be difficult or impossible to make a reliable empirical investigation of an unknown network of interpersonal choices or contacts if these are in some respect sensitive for the individuals involved. We consider how choice data given anonymously can be used to reveal various features of the network. Using tools from statistical decision theory and information theory we can determine risks and capacities of disclosing choice data, the graph structure of choice data and its numbers of loops and mutuals. We give some general results, illustrate the combinatorial complexity of the problem and comment upon the computational difficulties.     

Projective (2n,n,λ,1)-designs

A projective (2n,n,λ,1)-design is a set $\text{D}$ of n element subsets (called blocks) of a 2n-element set V having the properties that each element of V is a member of λ blocks and every two blocks have a non-empty intersection. This paper establishes existence and non-existence results for various projective (2n,n,λ,1)-designs and their subdesigns.