Tangential flow filtration (TFF) is a standard unit operation for concentration, buffer exchange and formulation across biopharmaceutical manufacturing. Process development effort is typically directed at membrane selection,
transmembrane pressure (TMP) operating point and crossflow rate. The pump is rarely treated as a variable. For the modalities that increasingly define the gene and cell therapy pipeline, including adeno-associated virus (AAV) vectors, lentiviral vectors, lipid nanoparticles (LNPs) and plasmid DNA (pDNA), this represents a systematic gap. TFF hardware was standardised around peristaltic architectures developed for robust proteins and over the last decade has been partially displaced by even-number positive-displacement diaphragm pumps, most prominently four-chamber configurations. Both architectures introduce mechanical environments (shear at the compression or chamber boundary, transmembrane pressure pulsation, suction-side transients at intensified flow) that are not neutral to structurally fragile products. This editorial argues that pump architecture should be evaluated as an independent, characterisable process parameter, that the literature gaps around it are themselves diagnostic and
that the regulatory framework already provides for this treatment under ICH Q8(R2). The aim is not to advocate a
specific architecture; it is to argue that the field has stopped asking what the pump is doing.