1. 2005.06.28. Dr. Pogány - WHO, Pretoria 1/59 Supplementary Training Workshop on Good Manufacturing Practices (GMP) MANUFACTURING PROCESS VALIDATION Solid Dosage Forms János Pogány, pharmacist, PhD, consultant to WHO Pretoria, South Africa, 28 June 2005 E-mail: pogany@t-online.hupogany@t-online.hu

2. 2005.06.28. Dr. Pogány - WHO, Pretoria 2/59 WHO GMP and related guides  WHO good manufacturing practices (GMP): main principles for pharmaceutical products  Section 4. Qualification and validation  Supplementary guidelines on good manufacturing practices (GMP): Validation (2003) – Draft.

3. 2005.06.28. Dr. Pogány - WHO, Pretoria 3/59 WHO GMP and related guides  WHO good manufacturing practices: main principles for pharmaceutical products – Validation of manufacturing processes  Good manufacturing practices for pharmaceutical products. In: WHO Expert Committee on Specifications for Pharmaceutical Preparations. Thirty-second report. Geneva, World Health Organization, 1992:14–79 (WHO Technical Report Series, No. 823).

4. 2005.06.28. Dr. Pogány - WHO, Pretoria 4/59 ICH guidelines  PHARMACEUTICAL DEVELOPMENT, Q8, Draft ICH Consensus Guideline, Released for Consultation on 18 November 2004, at Step 2 of the ICH Process  QUALITY RISK MANAGEMENT, Q9, Draft ICH Consensus Guideline, Released for Consultation on 22 March 2005, at Step 2 of the ICH Process

5. PROSPECTIVE VALIDATION Pharmaceutical Development Laboratory scale R + D

6. 2005.06.28. Dr. Pogány - WHO, Pretoria 6/59 Physicochemical and physical properties of API Physicochemical  hygroscopicity  solubility  water content  polymorphism  permeability Physical  particle size  bulk density (g/100ml)  flowability  color, olor, taste  consistency

7. 2005.06.28. Dr. Pogány - WHO, Pretoria 7/59 Equilibrium Moisture Content At relative humidities (RHs) <100%, a solid API (that does not form crystalline compounds with water) will loose some bound and all its unbound water until it is in equilibrium with the surrounding atmosphere. The sum of both these moistures is the free moisture of the API (granules) at the specified RH.

8. 2005.06.28. Dr. Pogány - WHO, Pretoria 8/59 Rate of Water Absorption at Different RHs

9. 2005.06.28. Dr. Pogány - WHO, Pretoria 9/59 Solubility of Zidovudine at 25 o C pHDissolved material (mg/ml) 3.021 4.020 5.320 6.021 7.022 Distilled water20 8.021

10. 2005.06.28. Dr. Pogány - WHO, Pretoria 10/59 Solubility of Artesunate pHDissolved API (mg/ml) 11,9 51,5 63,5 710,2 812,2

11. 2005.06.28. Dr. Pogány - WHO, Pretoria 11/59 Decomposition of Artesunate in aqueous solution SolventTime (h)Decomposition (%) Water20 0.1N HCl274 0.1N NaOH2100

12. 2005.06.28. Dr. Pogány - WHO, Pretoria 12/59 Relationship between Permeability Coefficient and Octanol-Water Partition 1 Prednisolone... 3 Dexamethazone... 9 Dexamethazone-acetate... 11 Progesterone

13. 2005.06.28. Dr. Pogány - WHO, Pretoria 13/59 NORVIR (Ritonavir) EPAR/CPMP /527/96 1.No polymorphism observed at the time of first submission (only form I : hard capsules and oral solution registered) 2.Failure in dissolution during stability studies for hard capsules 3.Emergence of form II (contamination of form I) 4.Production of hard capsules discontinued 5.Development and registration of soft capsules

14. 2005.06.28. Dr. Pogány - WHO, Pretoria 14/59 Particle Size When the solubility of an API is less than 0.1 mg/ml, the optimization of the particle size during preformulation may be critical to efficacy or pharmaceutical equivalence. Other researchers believe that particle size may be critical at a solubility of 1 mg/ml or less.

15. 2005.06.28. Dr. Pogány - WHO, Pretoria 15/59 Effect of Particle Size on Dissolution of Nevirapine tablets

16. 2005.06.28. Dr. Pogány - WHO, Pretoria 16/59 Screening of Compositions  Compatibility of an API with the excipients and the APIs with each other in FDCs is studied in open system stress stability experiments, e.g., 60- 80 o C, 100% RH.  Regulatory stability studies of the final composition are frequently initiated in the pharmaceutical R + D laboratory.

17. 2005.06.28. Dr. Pogány - WHO, Pretoria 17/59 Compatibility of Acetylsalicylic Acid with Excipients Time (week)Talc ATalc B Salicylic acid, % 00.10 40.325.85 80.4113.00 120.8028.50

18. 2005.06.28. Dr. Pogány - WHO, Pretoria 18/59 Triomune - WHOPAR Experimental „studies showed chemical incompatibility for the lamivudine with stavudine and nevirapine with stavudine combination. Lamivudine with nevirapine showed no change indicating that they are compatible. Stavudine was found incompatible with both the drugs, indicated by the brown colouration and increase in the impurities. Therefore it was decided to separate stavudine from the other two drugs. Hence the formulation was proposed to be bilayered tablet formulation, where stavudine is in one layer and lamivudine + nevirapine in other layer. Thus contact of stavudine with the other two drugs was minimised.”

19. 2005.06.28. Dr. Pogány - WHO, Pretoria 19/59 Dissolution Test and Profile  A (discriminating) dissolution test method should be developed for the final composition of the FPP.  Limits should be set for each API in fixed-dose FPPs.  The dissolution method should be incorporated into the stability and quality control programs.  Multipoint dissolution profiles of both the test and the reference FPPs should be compared.

20. 2005.06.28. Dr. Pogány - WHO, Pretoria 20/59 Dissolution Profile of Viramune and Generic Nevirapine Tablets on the Indian Market F2F2 20 73

21. 2005.06.28. Dr. Pogány - WHO, Pretoria 21/59 Hypothetical Dissolution Profile of a 2-FDC FPP

22. 2005.06.28. Dr. Pogány - WHO, Pretoria 22/59 Pivotal Batches A tabulated summary of the compositions of the clinical, bioequivalence, stability and validation FPP batches together with documentation (batch number, batch size, manufacturing date and certificate of analysis at batch release) and a presentation of dissolution profiles must be provided. Results from comparative in vitro studies (e.g., dissolution) or comparative in vivo studies (e.g., bioequivalence) should be discussed when appropriate.

23. 2005.06.28. Dr. Pogány - WHO, Pretoria 23/59 Excipients – Lactose (L) Different grade, different physical properties:  Angle of repose: 32- 47 o (Specs.)  Bulk density: 0.34 – 0.80 g/cm 3 (Specs.)  Bulk density (tapped): 0.41 – 0.95 g/cm 3  Flowability (spray processed): 4.1 g/s (Specs.)  Hygroscopicity : L monohydrate is stable in air at room temperature. Anhydrous L may absorb humidity.  Moisture content: L monohydrate contains approx. 5% w/w water of crystallization

24. 2005.06.28. Dr. Pogány - WHO, Pretoria 24/59 Excipients – Lactose (L) Solubility in water  1 in 4.63 at 25 o C  1 in 3.14 at 40 o C  1 in 2.04 at 50 o C  1 in 1.68 at 60 o C  1 in 1.07 at 80 o C Particle size distribution: depends on grade. Stability: may develop brown colouration (≥ 80% RH) Incompatibility : APIs with a primary amine group (base catalysed), aminophylline and amphetamines.

25. 2005.06.28. Dr. Pogány - WHO, Pretoria 25/59 Surface of a film-coated tablets containing a high level of a superdisintegrant

26. 2005.06.28. Dr. Pogány - WHO, Pretoria 26/59 Packaging Materials  Moisture-impermeable containers: glass ampoules, vials closed with rubber stoppers and fixed with metal caps, aluminium/aluminium blisters, high density polyethylene (HDPE) or glass bottles fitted with metal metal or HDPE closures, etc.  Moisture-permeable containers: polyvinyl chloride (PVC) blisters, low density polyethylene (LDPE) bottles, HDPE bottles fitted with polypropylene closures.  Specifications of packaging materials should include thickness and permeability coefficient.

27. CONCURRENT VALIDATION Commitment Batches

28. 2005.06.28. Dr. Pogány - WHO, Pretoria 28/59 Technical pharmacy  Pharmaceutical production system (from purchasing API to packaging FP)  Utility support system (HVAC, water, HPLC, etc. equipment containing many items)  Process (tablet making)  (Unit) operation (granulation, compression)  Step (sifting, sizing)  Procedure, method, technique (SOP)

29. 2005.06.28. Dr. Pogány - WHO, Pretoria 29/59 Causes of variation  Man (different operators - lack of proper training)  Machine / equipment (variation of tablet weight)  Measurement (lack of calibration)  Method (validated manufacturing methods)  Material (batch-to-batch variation of the same crystal form – different crystal forms (ASA)]  Environment (OoS T and RH in capsule filling)

30. 2005.06.28. Dr. Pogány - WHO, Pretoria 30/59 4.10 Scientific approach  „Processes and procedures should be established on the basis of the results of the validation performed.” Objectives  To prove that the tests, measurements, results and interpretation of formal studies on (manufacturing) processes and procedures/methods are appropriate and accurate.  To stabilize new processes (to reduce variability, to increase batch to batch consistency of quality attributes of products).  To reduce defect levels (standardize yields).  To reduce production costs.

31. 2005.06.28. Dr. Pogány - WHO, Pretoria 31/59 Process approach CONTINUOUS IMPROVEMENT OF THE QUALITY MANAGEMENT SYSTEM CUSTOMERCUSTOMER REQUIREMENTSREQUIREMENTS CUSTOMERCUSTOMER SATISFACTIONSATISFACTION Management responsibility Resource management Monitoring, improvement Manufacture Product Inputs

32. 2005.06.28. Dr. Pogány - WHO, Pretoria 32/59 Measure of variation (spread of data) 95.46% 68.26%

33. 2005.06.28. Dr. Pogány - WHO, Pretoria 33/59 Mean (average) chart Normal variation due to common causes UCL Upper control limit LCLLower control limit Abnormal variation of process – special causes average = mean

34. 2005.06.28. Dr. Pogány - WHO, Pretoria 34/59 Process capability index, Cp acceptance limitsUCL - LCL Cp = = process capability 6σ* 6σ* Three sigma: Cp = = 1 6σ* 12σ* Six sigma: Cp = = 2 6σ* σ*... is the measured standard deviation of the process

35. 2005.06.28. Dr. Pogány - WHO, Pretoria 35/59 Process capability index, Cpk UCL - x Cpk = 3σ n UCL... upper control limit x... mean of the acceptance criteria, target value σ n... is 50% of the measured standard deviation of the process Cpk shows the closeness of the process mean to the target value.

36. 2005.06.28. Dr. Pogány - WHO, Pretoria 36/59 Output of processes for different Cpk indices nσ n = 3σ UCLCpkNo. of products OoS (ppm) 1σ11σ1 m + σ 1 -0.166691 464 2σ22σ2 m + 2 σ 2 0.166308 536 3σ33σ3 m + 3 σ 3 0.566 807 4σ44σ4 m + 4 σ 4 0.8336 210 5σ55σ5 m + 5 σ 5 1.166233 6σ66σ6 m + 6 σ 6 1.53.4

37. 2005.06.28. Dr. Pogány - WHO, Pretoria 37/59 Objective and result of process control 1.The process reveals serious risks and it is not controlled 2.The process is not yet controlled but acceptance criteria are met 3.The process is under control and the product has a consistently high quality UCL N LCL UCL N LCL UCL N LCL

38. 2005.06.28. Dr. Pogány - WHO, Pretoria 38/59 Process under control  Most points fall near the central line (68% within one σ)  A few points fall near the control limits (5% in the third σ)  Points shold balance on both sides of the mean  Points should cross the mean line often.  Points should show a random pattern (no trends, cycles, clustering)

39. 2005.06.28. Dr. Pogány - WHO, Pretoria 39/59 4.8-4.9 Protocols and reports  Validation studies are an essential part of GMP and should be conducted in accordance with predefined and approved protocols.  A written report summarizing the results recorded and the conclusions reached should be prepared and stored.

40. 2005.06.28. Dr. Pogány - WHO, Pretoria 40/59 Process validation protocol/report  Short description of the process with a summary of the critical processing steps or critical parameters to be monitored during validation.  Additional testing intended to be carried out (e.g. with proposed acceptance criteria and analytical validation as appropriate).  Sampling plan — where, when, how and how many samples are taken.  Details of methods for recording and evaluation of results.

41. 2005.06.28. Dr. Pogány - WHO, Pretoria 41/59 Illustrative variables of wet granulation Process stepControl or manipulate (independent) variables Measured responses or output (dependent) variables Crystallization Micronization Particle size Bulk density Dissolution time Granulation and granule variables Pre-mixingSpeed, time, order of addition Blend uniformity Wet kneadingBatch (load) size Speed  Impeller  Chopper Spraying rate Volume of binder solution Granulation time End-point amperage  Impeller  Chopper Additional solvent volume

42. 2005.06.28. Dr. Pogány - WHO, Pretoria 42/59 Illustrative variables of wet granulation Process stepControl or manipulate (independent) variables Measured responses or output (dependent) variables DryingInlet air temperature (seasonal variation) Drying time (seasonal variation) Cooling time (if applicable) Outlet air temperature LOD Moisture content SizingScreen type and size Feed rate Granule size distribution (variation of sub-batches) BlendingBatch size (sub-batches) Speed Blending time Blend uniformity Bulk density  untapped  tapped Flowability Yield

43. 2005.06.28. Dr. Pogány - WHO, Pretoria 43/59 Indentation hardness profiles for tablets of different shape  Flat  Shallow convex  Standard convex  Deep convex  Ball-shaped

44. 2005.06.28. Dr. Pogány - WHO, Pretoria 44/59 Areas most prone to surface erosion for flat, shallow convex, caplet-shaped and deep convex tablets

45. 2005.06.28. Dr. Pogány - WHO, Pretoria 45/59 Erosion on the surface of the tablet with a logo

46. 2005.06.28. Dr. Pogány - WHO, Pretoria 46/59 Twinning during the coating process for flat-faced and caplet shaped tablets

47. 2005.06.28. Dr. Pogány - WHO, Pretoria 47/59 Measurement points of film thickness across the tablet surfaces FACE EDGE SIDE

48. 2005.06.28. Dr. Pogány - WHO, Pretoria 48/59 Illustrative variables of compression and film-coating Process stepControl or manipulate (independent) variables Measured responses or output (dependent) variables CompressionMachine speed Granule feed rate Precompression force Compression force Punches and dies Weight variation Content uniformity Friability Hardness Thickness Disintegration Dissolution time and profile Yield Film-coatingInlet air temperature Inlet air flow Spray rate Spray atomizing pressure Outlet air temperature Tablet-bed temperature Coat quality Yield

49. 2005.06.28. Dr. Pogány - WHO, Pretoria 49/59 Illustrative variables of tablet packaging Process stepControl or manipulate (independent) variables Measured responses or output (dependent) variables BlisteringMachine speed Machinability of blister material Forming temperature Forming pressure Sealing temperature Sealing pressure Leak testing Appearance Minimum information is legible Yield Bulk packing Tablet counter Incomplete tablets Machine speed Number of tablets Detection, counting Pilfer-proof Labeling Yield

50. 2005.06.28. Dr. Pogány - WHO, Pretoria 50/59 4.10 Scientific approach  Processes and procedures should be established on the basis of the results of the validation performed.

51. 2005.06.28. Dr. Pogány - WHO, Pretoria 51/59 Commitment (validation) batches  Process validation reports should be submitted in the application for prequalification.  Formal studies of production scale batches (not less than three) are required to identify the critical variables.  Provisional equipment control parameters and the corresponding in-process acceptance criteria must be deduced from the results of experiments with the validation batches.  Critical parameters are to be monitored, non-critical ones should be tested occasionally.

52. RETROSPECTIVE VALIDATION Annual Product Review

53. 2005.06.28. Dr. Pogány - WHO, Pretoria 53/59 Annual FPP quality review (1)  Starting materials used in the product, especially those from new sources.  Critical in-process controls and finished product results.  All batches that failed to meet established specification(s).  All critical deviations or non-conformances and related investigations.  All changes carried out to the processes or analytical methods.  Marketing Authorisation variations submitted, or granted, or refused, including those for third country dossiers.

54. 2005.06.28. Dr. Pogány - WHO, Pretoria 54/59 Annual FPP quality review (2)  Results of the stability monitoring programme.  All quality-related returns, complaints and recalls, including export only medicinal products.  Adequacy of previous corrective actions.  For new marketing authorisations, a review of post- marketing commitments.  A list of validated procedures and their revalidation dates.  A list of qualified equipment, support utility systems and their requalification dates, including calibration programmes.

55. 2005.06.28. Dr. Pogány - WHO, Pretoria 55/59 Case summary of 20 batches (1) StatisticsAv. wt. mgDissolution %Assay % Mean347,699,698,2 Median346,9100,097,5 SD5,22.52.2 Range22.310.08.8 Minimum337.095.695.0 Maximum359.3105.6103.8 Conf. level, 95%2.41.31.0 Accept. Crit.350±5%75%, 40'90-110

56. 2005.06.28. Dr. Pogány - WHO, Pretoria 56/59 Case summary of 20 batches (2) 1.Acceptance criteria for assay and dissolution rate are loose and should be tightened. 2.Potentially critical impurities are not tested. 3.IPC data are not included in the retrospective analysis of batch records.

57. 2005.06.28. Dr. Pogány - WHO, Pretoria 57/59 BEST PROCESS MINIMUM REQUIRED INPUT MAXIMUM OUTPUT AT NO COST TO SOCIETY (industrial safety, labour safety, internal and external environment protection)

58. 2005.06.28. Dr. Pogány - WHO, Pretoria 58/59 COSTS OF QUALITY Visible costs, e.g., waste and returned goods Hidden costs, e.g., wrong decisions, non-competitive manufacturing process, low yield, maintenance, idle machine time, workers attitude, etc.

59. 2005.06.28. Dr. Pogány - WHO, Pretoria 59/59 Main points again  Manufacturing methods are the same in the innovative and generic industries.  Pharmaceutical development is a major source of early identification of critical product and process parameters.  Validation batches should be tested extensively to establish preliminary/tentative IPC parameters.  Annual product review results in continuous improvement of products and processes.