Issue link: https://www.e-digitaleditions.com/i/1248960
28 Pharmaceutical Technology BIOLOGICS AND STERILE DRUG MANUFACTURING 2020 P h a r mTe c h . c o m Manufacturing able. As expected, initial feed turbidity impacted volumetric throughput of the clarification. Recov- eries represent filtration trials and don't account for the post-filtration buffer flush recovery operation. Based on the data produced at small scale using worst-case feed streams (higher cell density leading to higher turbidity), this filter was then scaled to a feed load of 50 L using a single 0.55 m 2 filter processed at the best performing flux. Data are show in Figure 3. The process scale runs (50 L) provided accept- able capacities, with the loading of > 90 L/m 2 . The actual cell densities were lower than the small-scale development runs, giving confidence that the over- all sizing was achieved on a worse-case cell density scenario. There was very little increase in pressure during the process—a 0.06 Bar increase on the depth filter (Clarisolve, MilliporeSigma, Burling- ton, MA) and a 0.05 Bar over the bioburden (0.22 µm) reduction filter. Across four large-scale runs, average recovery of > 70% of physical titer (p24%) and > 95% infec- tious titer (TU/mL) were achieved. Average of HCP clearance was also positively affected > 85%. For a complete recovery, f lush post-filtration is essential and becomes more important with this filter (Clarisolve 60HX filters, MilliporeSigma, Burlington, MA); the 0.55 m 2 pods (Clarisolve, MilliporeSigma, Burlington, MA) can add between 12–14 L extra volume and, therefore, be relevant for the overall yield. Higher feed turbidities can result in lower turbid- ity clearance, potentially impacting the bioburden filter. This can be confirmed by monitoring tur- bidity at increasing throughputs and analyzing the subsequent pools on V max against the 0.22-µm filter. Utilization of a depth filter (Clarisolve 60HX, MilliporeSigma, Burlington, MA) as the only clari- fication filter for a HEK-293T cell-culture produc- ing lentivirus is possible and was demonstrated at large scale, but it is evident that this will be subject to culture conditions. If pressure increases on the bioburden reduction filter become challenging, this can be addressed with use of a polishing sec- ondary clarification/protection filter beforehand (e.g., single-layered Millistak+ CE filter [CE40, CE45, or CE50], MilliporeSigma, Burlington, MA) or an inert polypropylene membrane prefilter (e.g., Polygard CN, MilliporeSigma, Burlington, MA). Figure 3. Representative pressure profile of a large- scale run with bioburden reduction filter in line. "Process" Scale pressure curve Volumetric Throughput (L/m 2 ) Bioburden reduction CS60HX (0.55 m 2 ) Pressure (bar) 0.14 70 80 90 100 60 50 40 30 20 10 0 0.12 0.08 0.1 0.06 0.04 0.02 0 Figure 2. Pressure resistance profiles with respect to throughput with a human embryonic kidney (HEK)- 293T cell-culture feed actively producing lentivirus product. LMH is liters per meter per hour. Resistance vs Throughout showing Turbidity reduction achieved Resistance (Bar/LMH) Throughput (L/m 2 ) Turbidity Reduction 0.010 140 120 100 80 60 40 20 0 0.008 0.006 0.004 0.002 0.000 Turbidity Reduction % 140 150 LMH 23 150 LMH 23 75 LMH 100% 99% 98% 97% 96% 95% 94% 93% 92% 91% 90%