Figure legends
Figure 1 Filterability of various column chromatography flow-through process outputs of aggregate-spiked mAb (1.0% aggregate spike) shown as (a) throughput against filtration time and (b) permeate flux (L/(m2 h)) against throughput. For all filtrations, mAb at 11 mg/mL in 20 mM Tris-Acetate, pH 7 adjusted to 15 mS/cm with NaCl was filtered at 0.3 MPa. Reference (no aggregate spike) and control (aggregate spike) have no column chromatography treatment.
Figure 2 Size exclusion chromatography profiles of mAb solutions. Profiles of mAb without aggregate spike (reference) and with 1.0% aggregate spike (control) are shown along with mixed-mode AEX1 flow-through fraction of aggregate-spiked mAb and nylon prefilter filtrate of aggregate-spiked mAb. Peaks of monomer, dimer and trimer and larger (≥Trimer) aggregates are labeled.
Figure 3 Filterability of various column chromatography flow-through process outputs of aggregate-spiked plasma IgG (0.5% aggregate spike) shown as (a) throughput against filtration time and (b) permeate flux (L/m2 h) against throughput. For all filtrations, plasma IgG at 11 mg/mL in 20 mM sodium acetate, 100 mM NaCl, pH 6 was filtered at 0.35 MPa. Reference (no aggregate spike) and control (aggregate spike) have no column chromatography treatment.
Figure 4 Comparison of clogging models for aggregate-spiked mAb (1.0% aggregate spike) filtrations shown in Figure 1. Experimental values and calculated values for each model are presented for each of these solutions: (a) reference (no aggregate spike), (b) control (1.0% aggregate spike), (c) mixed-mode AEX1 flow-through output of aggregate-spiked mAb, (d) normal AEX flow-through output of aggregate-spiked mAb. Clogging models: cake filtration, intermediate, standard and complete. For all filtrations, mAb at 11 mg/mL in 20 mM Tris-Acetate, pH 7 adjusted to 15 mS/cm with NaCl was filtered at 0.3 MPa.
Figure 5 Clogging model analysis for aggregate-spiked mAb (1.0% aggregate spike) filtrations shown in Figure 1 based on (a) clogging factor, k, and (b) mean difference between experimental values and calculated values for each clogging model (calculated using Equation 9). Clogging models: cake filtration, intermediate blocking, standard blocking and complete blocking. mAb at 11 mg/mL in 20 mM Tris-Acetate, pH 7 adjusted to 15 mS/cm with NaCl was filtered at 0.3 MPa. Reference (no aggregate spike) and control (aggregate spike) have no column chromatography treatment.
Figure 6 Comparison of clogging models for aggregate-spiked plasma IgG (0.5% IgG aggregate spike) filtrations shown in Figure 3. Experimental values and calculated values for each model are presented for each of these solutions: (a) reference (no aggregate spike), (b) control (0.5% aggregate spike), (c) modified CEX1 flow-through output of aggregate-spiked plasma IgG, (d) normal AEX flow-through output of aggregate-spiked plasma IgG. Clogging models: cake filtration, intermediate blocking, standard blocking and complete blocking. For all filtrations, plasma IgG at 11 mg/mL in 20 mM sodium acetate, 100 mM NaCl, pH 6 was filtered at 0.35 MPa.
Figure 7 Clogging model analysis for filtration of aggregate-spiked plasma IgG (0.5% aggregate spike) shown in Figure 3 based on (a) clogging factor, k, and (b) mean difference between experimental values and calculated values for each clogging model (calculated using Equation 9). Clogging models: cake filtration, intermediate blocking, standard blocking and complete blocking. Plasma IgG at 11 mg/mL in 20 mM sodium acetate, 100 mM NaCl, pH 6 was filtered at 0.35 MPa. Reference (no aggregate spike) and control (aggregate spike) have no column chromatography treatment.
Figure 8 Extrapolation of experimental filtration values of plasma IgG reference solution (no aggregate spike) shown in Figure 6a based on calculated values for each clogging model. Plasma IgG at 11 mg/mL in 20 mM sodium acetate, 100 mM NaCl, pH 6 was filtered at 0.35 MPa. Clogging models: cake filtration, intermediate blocking, standard blocking and complete blocking.