LAUSR

Dear Editors, We have read with interest the comments by Schmitt et al to our study of the “cost of decentralized CAR T-cell production in an academic setting.” While the letter confirms the sensitivity of the topic, in our view it is, at least in part, a result of misunderstandings by its authors. Importantly, our study is not based on “scientific modeling” (as suggested in the letter) but rather on the application of standard cost accounting techniques using empirical data. Our objective has been to increase the transparency of one cost component often cited as a reason for certain pricing policies, and to illuminate the potential to reduce manufacturing cost and improve efficiency offered by an alternative to commercial products and their high acquisition costs. Unfortunately, in their letter, the authors do not contribute to the much-needed transparency, as they claim that in their experience the (undefined) “real-life financial effort ... differs dramatically” from our analysis, but do not offer any quantitative information to support their claim. For clarification, let us first reiterate that our report deals with just one cost component, that is, the fixed and variable cost arising from decentralized chimeric antigen receptor (CAR) T-cell production in our setting. This should not be confused with pricing and reimbursement policies. Especially in commercial settings, the latter will also be influenced by the cost of research and development, by administrative overheads including marketing and sales expenditures, as well as by the expected return on investment. We note that “clinical development” (the last issue raised in the letter by Schmitt et al) represents a cost category clearly different from the cost of production. In the absence of transparent information provided in the letter, we cannot explain why its authors experience “dramatically different” and supposedly higher costs. Instead of speculating about potential underlying factors such as degree of process automation and use of modern closed systems, we would like to take this opportunity to explain our calculation by responding to the concerns raised by Schmitt et al in detail.