Validation of the BacT/Alert as an Alternative Sterility Test for Dendreon s Autologous Cell Therapy Product Timothy Wood QC Scientist Somatic Cell Therapy Symposium, Bethesda MD September 28, 2007 1
Overview of Manufacturing and Sterility Testing 2
Challenges of 21 CFR 610.12 (or USP Method) Product can render microbiological media turbid Insufficient time for an accurate go no go pre-harvest read Limited shelf life: < 24 hours Sub-culture transfer if necessary would occur post-infusion No bulk product. The Culture Pool (in-process) sterility has been tested to the CFR bulk Repeat testing rules are not appropriate for our product 3
Method Validation Challenges Experimental Design No standard available Guidance obtained from the new PDA TR # 33 (Parenteral Drug Assoc.) Guidance from draft USP <1223> Validation of Alternative Micro Methods Input from industry colleagues experience Equivalence challenges Fresh samples from donors for seeded studies are limited and can be logistically challenging Non-seeded side by side testing not possible (limited dose) How to demonstrate when the standard (21 CFR 610.12) does not provide needed assurance? 4
Key reasons for selection of BacT/Alert method as alternative Non-destructive (ability to directly subculture positives for ID) Faster detection Sample volume (representative) No sample filtration needed Minimal preparation and manipulation (reduces false positives) Continuous monitoring Similarities to CFR standard (aerobic, anaerobic, sample to media ratio) FDA approved for blood cultures and platelet sterility testing Testing throughput 5
BacT/Alert Validation Plan Outline (Phase I & III Products) Installation Qualification (IQ) Feasibility (Proof of Concept) Studies -System suitability (detection of USP panel organisms) -Comparison study with CFR method using fresh healthy donor product samples Operational Qualification (OQ) -Verification of BacT/Alert system operations -Temperature mapping of incubation chambers 6
Validation Plan Outline (continued) Method Validation Study Design -6 ATCC strains (USP <71> panel) Staphylococcus aureus ATCC 6538 Bacillus subtilis ATCC6633 Pseudomonas aeruginosaatcc9027 Clostridium sporogenes ATCC 11437 Candida albicans ATCC 10231 Aspergillus niger ATCC 16404-6 environmental / product isolates Bacillus dipsosauri Staphylococcus epidermidis Brevibacillus brevi Micrococcus luteus Rhodococcus sp. Bacillus sp. 7
Validation Plan Outline (continued) Method Validation Study -Test article in triplicates vs. non-product control bottles without product (B/F -bacteriostasis/fungistasis) -BacT/Alert Incubation: 35 C for up to 14 days -Acceptance criteria: -Specificity (ATCC strains & isolates detected in products) -Detection Time (< 7 days) -LOD (Sensitivity <10 CFU/bottle) -Repeatability/Ruggedness (multiple product lots, culture bottles, operators) 8
Feasibility: Suitability of BacT/Alert Culture bottles Hrs 96 72 48 24 0 Growth Promotion < 100 CFU Time to Positive BT/A bottles vs CFR/USP Method (typical values) BacT/Alert USP Method B subtilis 6633 S aureus 6538 P ae ruginosa 9027 C sporogenes 11437 C albicans 10231 A niger 16404 ATCC strain 9
Feasibility: BacT/Alert vscfr (Culture Pool) Positive Detection Times (3 replicates) Hrs 60 50 40 30 20 10 0 1/3 3/3 0/3 0/3 3/3 2/3 3/3 3/3 3/3 3/3 1/3 1/3 BTA Method CFR (FTM media) B. subtilis 46 CFU B. subtilis 6 CFU B. subtilis 2 CFU S. aureus 65 CFU S. aureus 8 CFU S. aureus 1 CFU CFR: 10 ml sample into 100mL TSB (20-25 C) and 100 mlftm (30-35 C) BacT/Alert: 5 ml sample into 40 ml aerobic and anaerobic bottles (35 C) 10
Feasibility: BacT/Alert vscfr (Final Product) Positive Detection Times (2 replicates) Hrs 100 80 60 40 20 0 2/2 2/2 2/2 2/2 2/2 2/2 0/2 1/2 1/2 BTA Method CFR (FTM) CFR (TSB) S. a ure us 69 CFU S. a ure us 7 CFU S. a ure us 1 CFU 11
Feasibility Study Conclusions Cell product is compatible with the BacT/Alert method Suitability (growth promotion) of culture bottles confirmed Overall BacT/Alert performed better than CFR CFR false negative reads (visual): 4 BacT/Alert false negatives: 0 Detection time range BacT/Alert : 12.1-18.4 hrs CFR: 24 hrs not detected (14 days) 12
Method Validation: B/F Study of ATCC Panel <100 CFU Final product vsnon-product control bottles (35 C) Avg Hrs to Positive (n=3) 60 50 40 30 20 10 0 12.3 12.9 13.9 20.3 B subtilis 43 CFU S a ureu s 53 CFU (AST) 15.8 18.5 16.7 18.4 S aureu s 53 CFU (NST) P a erug inosa 52 CFU C albicans 2 3 CFU 25.1 27 51 20.3 C sporo genes 16 CFU (NST) A n iger 25 CFU 39.9 44 NEG (14 days) Neg Controls 0 CFU Cell Product Pos Control 13
Method Validation: Environmental & Product Isolates Final product vsnon-product control bottles (35 C) 100 91.7 H rs to Positive (n=3) 80 60 40 20 0 11.1 11.3 23.9 24 25.9 26.1 28.1 46.6 47.6 24.9 73.7 31.2 33.5 NEG (14 days ) Cell Product Pos Control Bacillus dipsosauri 24 CFU Microco ccus luteus 73 CFU Sta ph epiderm idis 37 CF U (AS T) Sta ph epiderm idis 37 CFU (NST) Rhodococcus spp 103 CF U Brev ibac illus brevis 68 CFU Bacillus sp p 7 CFU Neg Cont rols 0 CFU 14
Method Validation: Sensitivity of Detection (LOD) Avg Hrs to Positive Detection 70 60 50 40 30 20 10 0 14.7 B. su bt ilis 6 CFU 20.0 S. aur eus 4 CFU (AST) 20.9 28.3 P. aeruginosa 5 CFU S. au r eus 4 CFU (NST) 29.0 C. a lbicans 8 CFU Target < 10 CFU/bottle 22.7 C. sporogen es 1 CFU (NST) 51.7 A. n iger 1 0 CFU 12.4 Bacillus d ip so sauri 5 CFU 30.3 Micr ococcu s lut eus 5 CFU 31.4 St ap h epiderm idis 5 CFU (AST) 21.5 St aph epiderm idis 5 CFU (NST) ATCC Strains and Isolates 60 Rhodococcus s p. 3 CFU 53.7 Breviba cillu s brevis 4 CFU NEG Bacillus sp. 2 CFU 15
Method Validation Initial Conclusions B/F studies concluded no effect of the product on detection All unseeded product controls were determined negative Range of average detection time (<100 CFU in product) 9.4 hrs Bacillus dipsosauri (isolate) 51 hrs Clostridium sporogenes (ATCC) Sensitivity (LOD) Range 1-10 CFU/culture bottle Repeatability/Ruggedness Reproducible results between multiple products, culture bottle lots, and operators 16
Further studies were generated following correspondence with FDA on validation progress Address slow-growing/low CO2 producing bacteria Include another mold Other microorganisms selected: Penicillium chrysogenum (low temperature mold, slow-growing) Propionibacteriumacnes (slow-growing) Pseudomonas fluorescens (associated with contaminated blood products) 17
Other Microorganisms Final product vsnon-product control bottles (35 C) Final Product Sample A v g H rs to Po s itive (n =3 ) 120 100 80 60 40 20 0 18.419.3 20.321.6 83 90 91 101 NEG (14 days) NEG (14 days) NEG (14 days) Cell Product Pos Control P fluor escens 49838 >200 CFU P fluor escens 49838 26 CFU P acnes 11827 90 CFU P acnes 11827 10 CFU P ch r ysogen um 10106 37 CFU P ch r ysogen um 10106 4 CFU Penicillium chyrsogenum: No growth or detection at 35 C Neg Contr ols 0 CFU 18
CO 2 production and detection of P. chrysogenum at lower incubation temperatures (standard algorithm) Penicillium chrysogenum incubation at 32 C CFU count = 29 Penicillium chrysogenum incubation at 28 C CFU count = 10 19
Method Validation Study Final Conclusions Wide range of organisms detected in products (<100 CFU) Products are compatible (0 false positives) Low-level sensitivity: 1-10 CFU/bottle Longest seeded detection time noted: Propionibacterium acnes 3 CFU @ 129 hours (5.4 days) Lower temperature increased sensitivity for Penicillium chrysogenum 20
Proposed BacT/Alert Sterility Method Feb 2006: BacT/Alert 7-day sterility method approved by FDA as an alternate sterility test for autologous Phase III product 21
Additional Data Requested by FDA Equivalence of BacT/Alert vs. CFR/USP Show comparison of product B/F results using panel of ATCC organisms Manufacturer s literature (aerobic & anaerobic bottles vs. TSB & FTM media) Comparability of detection algorithms (BacT/Alert Classic vs3d Models) 22
Equivalence Data: B/F Comparison between CFR and BacT/Alert (Culture Pool) 23
Equivalence Data: Manufacturer s literature 24
BacT/Alert Detection Comparability (with standard algorithm) Classic Model (Site 1-2002) vs 3D Model (Site 2-2006) Culture pool Final Product Microorganism & ATCC Number (<100 CFU) BT/A Bottle type BT/A Classic Model 120 Site 1 (2002) BT/A 3D Incubator 1 Site 2 (2006) BT/A 3D Incubator 2 Site 2 (2006) BT/A Classic Model 120 Site 1 (2002) BT/A 3D Incubator 1 Site 2 (2006) BT/A 3D Incubator 2 Site 2 (2006) B. subtilis 6633 aerobic (iast) 12.6 12.8 12.6 12.3 12.8 12.1 S. aureus 6538 aerobic (iast) 14.4 15.8 14.2 13.9 16.1 14.5 S. aureus 6538 anaerobic (inst) 15.6 16.3 15.5 15.8 16.6 15.8 P. aeruginosa 9027 aerobic (iast) 17.7 19.1 17.8 16.7 18.8 17.9 C. albicans 10231 aerobic (iast) 25.2 24.6 24.6 25.1 27.4 29.4 C. sporogenes 11437 anaerobic (inst) 30.9 22.1 20.2 51 19.7 20.5 A. niger 16404 aerobic (iast) 40.7 32.5 33.6 39.9 44.0 45.0 25
Moving Forward Established BacT/Alert validation requirements for new installations / new processing facilities Instrument IQ/OQ/PQ Performed local B/F study Included local isolates (to be licensed facility) Replace Gram stain release assay with rapid method Real-time results ( 4 hrs) Sensitivity 10 3 cfu Some possible technologies PCR (Total viable bacteria / yeast & mold) Fluorescent cytometry PGD Test (Verax Biomedical) BacSTAT (GenPrime) 26