Pharmaceutical Cleaning

This article is a review of the presentations made at a recent Seminar of the Pharmaceutical Committee of the Institution of Mechanical Engineers.  Over 100 delegates attended, several from Europe and one from as far away as Japan.

Speaker’s contributions were led by a presentation from an inspector of the regulatory body, the Medicines Control Agency.  Six suppliers with different approaches showed the breadth of the topic and two operators provided valuable input.

The Objectives of the Seminar

Cleaning issues are of huge importance in the Pharmaceutical & Biotechnology Industries.  The Regulator demands that companies must thoroughly assess the cleaning requirements of their processes, with the onus being on the companies and their suppliers to strike the right balance between regulatory compliance and techno/commercial constraints.

Introduction

You may think that clean in the pharmaceutical sector means the cleanest of industrial regimes.  Well, as the title of the seminar inferred, “To Clean or Too Clean”, you can do too much and just enough to meet the product and regulatory needs.

Pharmaceutical plants are certainly cleaner than food.  In the main, this is because you cannot “cook” the pill at 100ºC, to rid it of a few residual contaminants.  As one of the seminar planners remarked:

“Water intended for injection into pills can be classed as ‘almost muddy’ when compared with the quality needed in a microchip production facility”.

So the topic presented was intended to provoke thought on how clean was clean enough for pharmaceutical facilities. 

This seminar provided a refreshing review of the principles of cleaning technologies and considered the challenges faced by the end users and system designers in reaching the most technically feasible and cost-effective solutions. 

What were the Main Points?

The seminar concentrated on the main operational problem of Cleaning in Place (CIP).  It is one of the most important factors in the consistent quality of any pharmaceutical product.  Since most pharmaceutical products are produced in batches, it also should almost invariable be regarded as the final step in what is always a batch production cycle.

Apart from CIP, which requires a design to allow chemical cleaning and subsequent purging without dismantling, what other options are there? 

Well not surprisingly, cleaning out of place (COP), involving a manual strip-down of parts and components and then the tricky bit of their re-assembly without allowing the introduction of any contamination.  The same seminar planner added:

“The worst contaminants in any facility are people.”

There is also SIP, the sterilisation of the system with all its components in place, which like CIP requires the system’s design to have facilitating features.

And finally, of course, some combination of these techniques used as routine, or occasionally, special cleaning processes.

The basic variables of any of the processes are: -

  • ·             The chemical which is used
  • ·             The temperature at which it is applied
  • ·             The period of time it is applied
  • ·            

The soiling product determines if acid or alkali cleaning agents are to be used.  Either way they are hazardous to handle both before and after use and require an environmentally friendly disposal method.

As a rule of thumb an increase in temperature of 10ºC will improve the cleaning efficiency by 50%.

Time is an area of some experimentation as it brings in all the other variables that can exist.  The type of equipment and the process that has just ended.  The cleaning solution, its concentration and the temperature of use.

The wetting of the surfaces brings in the need for turbulent flow in the delivery pipes and the optimum placement of effective delivery nozzles, but more on that later.

Typical Processes

So a typical process built up by expert input with an overlay of trial and error would involve.

  1. A pre-rinse of fresh, warm water for five minutes with the water running to some subsequent de-contamination process.
  2. A hot alkali at 2% for 15 minutes and this agent can be re-circulated through a purpose designed recovery system.
  3. A further warm water rinse for 5 minutes and more water to the effluent treatment plant.
  4. A hot acid rinse at 1% for 10 minutes and then on to the acid recovery process.
  5. A final warm water rinse, but for 30 minutes, greatly adding to the effluent treatment load.
  6. And then a good old blow dry with sterile air going off to a simple air effluent emission process.

As you may guess there are many variants on this basic recipe.  They determine the types of delivery system, whether fixed or mobile, that may be employed.  The delivery temperatures, up to boiling and the extent of solvent recovery and reuse all impact the cleaning process equipment very heavily. 

Thus the cost effective cleaning process is a complex equation between the capital cost of the cleaning equipment, the solvent and water costs and acceptable load on the site effluent treatment facility.  Overriding all this is operator safety and the acceptable quality of the subsequent product production run.  Meeting all of these variables within acceptable regulatory and economic criteria often results in complex systems and is a job for a specialist at the outset of the product process design and cannot be done effectively as an afterthought.

How well does the Industry perform?

An MCA inspector expressed the views of the inspectorate on the interpretation of the relevant regulations.  These emphasise the need to design the equipment for cleaning from the outset.  Often insufficient attention was applied to the surface finish and the means by which it was obtained.

This also brought in the commissioning factors of ridding the process of a range of unknown contaminants created during the manufacturing and construction phases.  The inspector’s view was that these were usually well addressed during the initial plant start-up (IQ/OQ in pharmaceutical validation speak), but were not well handled after a major repair or shutdown.

Manual cleaning was also generally badly done.  Mops were often dirty, un-sterilised, and even rusty and the motivation of cleaners was obviously an unpredictable variant when compared with a mechanised process.

Mechanised systems could not be given a clean bill of health as few systems could detect if one of the nozzle outlets, so carefully proven in experiments was blocked.  They were also quite complicated processes in themselves, with many potential mal-functions unless they were closely monitored by manual or automated means.

That said, the Inspectorate felt that limits of residual contaminants at around one thousandth of the normal therapeutic dose was logical, practical and achievable.  It could also be verified and documented for subsequent inspections.

Overall, the increasing potency of drugs demand that the culture of URS, FATs and on site qualification be applied equally stringently to the cleaning process equipment as it is to the production process itself.

An interesting new development is that of disposable manufacturing “tanks”.  Whilst not applicable to all processes, the use in batch production of sterile, collapsible, plastic containers is being tested at several sites.

Further papers dealt with the chemistry of the reactive cleaning agents and how they could be made more effective, whether this is in higher strengths or reduced quantities, in order to reduce the amount of effluent to be disposed of. 

The use of solvents for cleaning primary manufacturing facilities is a common practice, but it is one that does not sit very comfortably with the regulators, production co-ordinators or accountants.  Whilst there are few residue issues with a solvent based cleaning system, they are typically found to be inefficient and frequently result in a “clean it until it’s clean” policy.  This usually manifests itself in the need to repeat “solvent boil-outs” until the process residues have been removed and this has been known to take as long as 16 refluxes.  When you consider that a single boil-up could take perhaps 2 days, it is sometimes observed that cleaning takes longer than manufacturing, especially in non-dedicated plants.  This “solvent abuse” increases the cost of the cleaning agents and energy costs, but perhaps more importantly, it seriously extends the cleaning time.  The knock-on effects of this are that plant efficiency is not as high as it could be and scheduling of the next batch/campaign can be delayed.

Comments that have been recently received include the ”loss of £2m in lost manufacturing time due to extended cleaning” or “our plant costs us £100,000 per day, so we can’t afford to spend a long time cleaning”. 

An alternative approach is the use of aqueous cleaning agents, and, providing the appropriate technologies and chemistries are applied, the rewards can be surprisingly large. 

The delivery of the cleaning agent to the surface in an even distribution produces some interesting experiments and theoretical analyses in fluid flow mechanics.  The latest rotary jet head technology should allow pharmaceutical manufacturers to optimise and streamline vessel-cleaning practices. 

Implementation of this technology will enable end users to save up to 50% or more on cleaning media consumption, reducing their environmental impact and making a significant reduction in process vessel downtime.  In certain cases, the vessel cleaning time can be reduced to a matter of hours rather than a procedure that normally takes days or even weeks.

The cleaning process can be modelled by means of simulation software, ensuring that all of the key criteria and configurations are met, leading to coverage guarantees prior to vessel fabrication.

The latest range of rotary jet heads is designed not only to clean the vessel effectively but also to self clean internally and externally, ensuring no product residues or microbial issues are present after the vessel wash cycle.

The validation system comprises of a pressure sensor that records the jet impacts within the vessel giving a continuous read out to the clean in place PLC control system.

Conclusions

Cleaning has been one of those subjects that, until 5 or 10 years ago, was too frequently considered of minor importance.  After all, cleaning does not generate revenue per se.  However, through a few instances highlighted by regulators, it has really been put in the spotlight.  Cleaning and cleaning validation is currently one of the hot topics for the regulators.

Throughout the papers the authors laid great stress on both the cleaning effectiveness of their views and the cost effectiveness of their suggestions.  This reflects the ever increasing importance that cleaning in the pharmaceutical sector is carried out to the standards required and at the costs the business can afford.

Unquestionably, the cost implications of cleaning a pharmaceutical facility are significant and in the primary pharmaceutical manufacturing environment there is scope for improvements on a number of levels.  These include the enhancement of the validation position, reduction in the amount of solvents and an improvement on plant capacity.  Naturally, the latter two points have a direct impact on the ability of the plant to operate in a more cost effective manner.

 

Norman Harris is an independent consultant specialising in Engineering Information Management Strategies and is Chair of the Pharmaceutical Committee of the Institution of Mechanical Engineers.

Queries can be addressed to consult@20cc.co.uk or visit the websites www.20cc.co.uk or www.imeche.org.uk for further information