Advancements in continuous tablet coating Jim Marjeram O Hara Technologies 16 October 2008 0
Agenda Background on tablet coating Motivation for continuous processing Current continuous coating capability Pharma requirements for continuous coating Advancements in continuous coating What the future holds? Discussion and questions 16 October 2008 1
Why coat tablets? Easier to swallow Masking taste and unpleasant smell Reducing light sensitivity and atmospheric oxidation improving stability Core is inelegant Colour migration from core Reduced friction in packaging process Modified drug release profile Separate incompatible substances 16 October 2008 2
Background on coating Tablet coating developed originally from the use of sugar to mask taste and provide attractive appearance to the core Tablets coated in batches in solid-wall pans (no perforations) Half a century ago tablet coating was restricted to sugar and solvent film based formulations by ability to remove moisture from coating 16 October 2008 3
Introduction of the sidevented tablet coating pans (with perforations) This evolution was required for the introduction of aqueous based film coating polymers to the pharma industry Carbon steel construction except for pan Many screws, not welded in places Not very GMP Last 40 years 16 October 2008 4
Last 25 years Introduction of reliable microprocessor based process control systems required to insure process control and repeatability Improved design spray nozzles for tablet coating specific applications (all stainless steel) Improved air preparation systems required for consistent aqueous process drying Improved GMP coater design, more cleanable, all stainless steel Improved tablet handling All required for the optimization of aqueous film coating process 16 October 2008 5
More advanced automatic CIP and WIP washing systems Last 15 years 16 October 2008 6
Last 5 years More advanced film coating spray nozzles with anti-bearding designs More reliable industrial automation for accurate and repeatable control of process parameters ie: dewpoint, mass solution flow, air flow etc. The evolution of the improvements to the batch tablet coater have allowed the recent advancements in continuous tablet coating 16 October 2008 7
Why continuous processing? To introduce process analytical technology to increase quality and reduce waste Quicker, easier and less costly scale up, all done in the same kit Ability to vary batch sizes based on product and demand Increased equipment utilization, less downtime Contained processes, start to finish, for potent products, less cleaning Lower manufacturing costs by many efficiencies Reduction in product losses, higher quality, reduction in variability, 6 sigma Reduction in capitol cost, smaller and more flexible machines Reduction in labour cost through high level of automation and more intense processes Reduction in space requirements, less GMP space Reduction in utility usage, cost savings Reduction in peak utility consumption, smaller plant Reduction in inventory, more flexible continuous lines Reduction in processing time, no storage, wait and material transport times 16 October 2008 8
Continuous solid dosage line Model courtesy of GSK 16 October 2008 9
Continuous coating intro 15 years ago introduction of the side-vented continuous tablet coating pans originally used and proven in the film seed coating industry Initially used in the nutritional supplements market in the US, later a little in the OTC Wider usage restricted by GMP requirements, pharma industry acceptance, large volumes required 16 October 2008 10
Continuous coating progress In the last 10 years more GMP pharma acceptable designs have been introduced, and designs are still progressing 16 October 2008 11
How does it work? Continuous coating pan is a stretching of the batch pan and has openings on front and back, or loading and discharge ends Spray manifolds are also stretched to the length of the coater Tablets are continuously gravity fed in at a very controlled rate at the charging end of coater via weightfeeder Average weight gain of tablets, or coating thickness is precisely controlled by tablet flow rate vs. solution flow rate, times efficiency Residence amount, residence time, pan RPM, bed depth, are all independent of tablet weight gain 16 October 2008 12
How does it work? The tablets move through the coater by natural migration, no baffles or vanes are used, unit is not sloped, simply the quicker the tablets are fed the quicker they move through, the lower the residence time Tablets are discharged at the unloading end of the coater by spilling around closure plate or weir By controlling the position of the weir, the level of tablets or residence amount of the coater can be controlled Adjustments to tablet feed rate, pan RPM and residence amount and/or time all effect the tablet uniformity 16 October 2008 13
Tablet cooling, elevating and/or waxing can be added at the discharge end of the coater How does it work? 16 October 2008 14
Further advancements for pharma Improved spray manifolds for even solution distribution Improved air caps to provide anti-bearding qualities required for continuous operation 16 October 2008 15
Further advancements Zero waste operation, allows every tablet to get coated Batch start up first step is to completely fill coater to predetermined level or weight and all nozzles start to spray just as a batch coater, discharge gate remains closed and tablet feeder is turned off. Nozzles begin to sequence off starting at the charging end at precalculated times This stages tablets with the appropriate weight gain at appropriate positions within the pan. 16 October 2008 16
Further advancements Zero waste operation batch start up progresses into continuous mode Tablet feeder starts and discharge weir opens to predetermined level Coater runs continuously until the last tablet from the hopper drops into pan, 100-30,000kg 16 October 2008 17
Further advancements Zero waste continuous operation progresses into batch shut down mode Tablet feeder stops and discharge weir closes All nozzles spray as start up initially but start to turn off in sequence starting at the discharge end until all the tablets are coated to the proper weight gain Spray stops, weir opens remaining tablets spill from the pan into the discharge chute 16 October 2008 18
Further advancements Batch start up coating capability has added benefit of being able to completely coat tablet batch without entering continuous mode Also 30 development pan can be used to precisely determine process parameters for 30 diameter stretched continuous pan at 1/10 th scale 16 October 2008 19
Further advancements for pharma Tablet diffusion testing and axial flow uniformity, tablets that go in must come out in similar order 16 October 2008 20
Initial Test Targets Initial target parameters for Continuous coating diffusion effects: Drum speed affects number of decisions or axial movements in a run as well as actual coating performance and is expected to have a definable relationship similar to a random walk (distance/square root of time) Bed depth may contribute to diffusion. i.e. the only tablet movement required is that past the nozzle and subsequent drying towards the bottom of the bed. Feed rate as a relationship to drum length for a given set of coating parameters. Higher atomization pressures will likely cause higher diffusion due to greater choice potential as the tablets pass the hole in the bed. Bed tilt may affect the speed of the tablets through the drum. Initial testing with no solution. Solution impact to be tested after main effects. Inlet conditions can affect overall residence time. 16 October 2008 21
Measuring Methods Multiple web cameras with multi-pc interface and custom software for analysis of video with marker tablets. Reasonably low cost solution to confirm exit conditions, compare start with exit performance, and increase accuracy of tests. 100 colored marker tablets with white tablet bed background started just ahead of spray zone for each set of conditions. 8 cameras running at 5 frames/second in the first half of the drum. Large quantities of raw data (200+ GB) Custom analysis software to look at pixel counts in each frame (post processed). Individual frame analysis to get information to improve exit data. (accurate start times and process lengths) Manual timing of marker tablet exit timing 2 people sorting through exiting tablets to retrieve and expose marker tablets 1 Time recorder attempting to record time that the marker tablets exit in relation to the test start time with use of small Excel macro. Models and tools DOE to see main effects Excel model to help predict future performance Monte Carlo simulation model to tease out relationships 16 October 2008 22
Lighting and Cameras on Spray Boom 16 October 2008 23
Lighting and Cameras on Boom 16 October 2008 24
Image Processing Images showing recognition (in green) of red marker tablets 16 October 2008 25
SD/Rev Testing Results Design-Expert Software SD/Rev 0.2137 0.0492 X1 = A: Feed Rate X2 = C: Bed Depth WORSE Uniformity 0.18 0.15 DOE shows Feed Rate and Bed Depth as major effects on tablet diffusion. Tilt affects bed depth. Actual Factors B: Tilt = 1.50 D: Atomazation = 0.50 0.12 0.09 BETTER 0.06 Uniformity 6 5.625 C: Bed Depth 5.25 4.875 4.5 600 750 900 1050 A: Feed Rate 1200 16 October 2008 26
SD/square root of revolutions (min/rev) Testing Results Timed results after enhancement for start times and positions from video data show how transport speed affects SD/square root of revolutions relationship that drives diffusion during each revolution of the drum. Tests done prior had estimated axial transport speed of <.01 m/rev with associated higher dispersion. 0.25 0.2 y = 0.0055x -0.8079 R 2 = 0.9435 0.15 0.1 0.05 0 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 Axial transport speed (m/rev) 16 October 2008 27
SD/sqrt Revolutions (min/rev) Testing Results Video results (in blue) scaled the same as and compared to the timed results (in red, same data as previous slide) show the same story. Variation in the video increased due to the early capture time and video intensity. 0.50 0.45 0.40 0.35 y = 0.0032x -0.9846 R 2 = 0.7094 0.30 0.25 y = 0.0055x -0.8079 R 2 = 0.9435 0.20 0.15 0.10 0.05 0.00 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 Axial Transport Speed (m/rev) 16 October 2008 28
Conclusions The relationship of bed depth and feed rate (axial transport speed) is the major driver for tablet diffusion/distribution growth rate and the resulting coating uniformity performance at a particular drum speed. Impact of atomization and pattern air is a visible but minor impact on uniformity performance. The variation of time in coating zone is what affects the tablet to tablet uniformity, not total residence time. (Allows mixing at ends for batch mode start up) Bed tilt has an indirect effect on diffusion, it is needed to maintain flat/uniform depth at specified feed rate. Bed tilt needs to change from batch to continuous. 16 October 2008 29
Backup Notes o Axial mixing efficiency controls uniformity of processing time. Minimum axial mixing creates more uniform residence time. Speed through the drum is the major influence. Drum rotation during processing should be as slow as possible. Actual impact of drum rotation speed has yet to be evaluated, but since drum speed affects number of decisions in a run it is anticipated that this will increase dispersion as well. Data suggests that at a constant drum rotation, the speed through the drum (m/min) and SD/rev are related. The feed in cubic meters/hr (not kg/hr) divided by the cross section of the bed/60 should equal the feed in meters per minute along the drum. The cross section of the drum at the target areas of interest 4-6.5 bed depth in the 30 diameter drum (100mm ~.036m^2 vs. 165mm at ~.075m^2) are about a 2X range Area of the cord segment (bed cross section) is: or C7^2*ATAN(SQRT(((C7/C9)^2)-1))-C9*SQRT(C7^2-C9^2)Where R or C7=pan radius/apothem and r or C9=pan radius-bed depth/sagitta 16 October 2008 30
Further advancements for pharma Proving coating uniformity equal or better than current batch coating standards by means of colour uniformity testing 16 October 2008 31
Colour uniformity testing Color development and uniformity testing. The pan load for the batch trial was 250kg. In the batch mode trial, samples were taken every minute and in continuous mode, samples were taken every 10 minutes for the duration of the trial and every minute during the tablet unloading cycle. These samples were tested instrumentally for color development and uniformity testing using a Diano Color Products Milton Roy Colormate employing the Commission Internationale de l Eclairage (CIE) L* a* b* system. Color-spectrophotometer Total color difference from target reference was determined by calculating the distance between two points in the color space using the following equation: ΔE* = [(L*1 L*2)2 + (a*1 a*2)2 + (b*1 b*2)2]1/2 The standard deviation of color difference between calculated ΔE values of the individual tablets from each set of samples were compared as a measure of coating uniformity. A ΔE value of <2 indicates no visual difference in color from the target reference color. 16 mins is 3% weight gain HPMC based Opadry II 16 October 2008 32
16 October 2008 33
16 October 2008 34
Backup O Hara coater Inside O Hara continuous coater during testing 16 October 2008 35
Near future Shrink continuous coating process down to capacity that matches that of most tablet presses, 10-150 kg/hr to better suit the needs of most pharma applications Shrunken single continuous coating unit could be used for R&D, Pilot and production coating Introduction of PAT to measure coating performance online 16 October 2008 36