Thursday, September 27, 2012

Patching Windows (OSI PI) Servers

A word to the wise:

When you're patching a Windows Server on which an OSIsoft PI server is installed, this is your worst enemy.

Restart Now

It sucks... big time, and here's why:

When you click Restart Now, it shuts down in a manner that corrupts the PI Event Queue. You're going to have data flowing to the PI Snapshot but doesn't make it into the PI Archives.

Then you have to call OSI TechSupport to dig yourself out of this mess.

Final message? Pay Attention! Click Close and go get yourself some coffee. You saved yourself 2-hours of hassle.

Get a PI Professional

Tuesday, September 25, 2012

E.Coli Replicating Near Thermodynamic Limits

Here's a fascinating Nature article translating a wonkish MIT paper on how E.coli are hyper efficient at replication.  And that their doubling time of 20 minutes is near the thermodynamic limit.

In all, it's an indecipherable 3-page thought experiment applying equations derived from cited sources, which - no doubt - apply equations derived from further sources.

With regard to cell culture, we already know that the 20-minute doubling time of microbes is the rule and that strange strange contaminants like Leptospira are the exception. 

Monday, September 24, 2012

@CellCultureDish On Eliminating Animal-Derived Components - ಠ_ಠ

In a recent article on key improvements for improving the quality of biopharmaceutical manufacturing, the author, "The Dish" states:

Many problems with respect to contamination are a result of problems with raw materials and too often these raw materials are sourced from animals. Animal-derived products always carry a risk of contamination from adventitious agents, such as viruses and prions. Not only are animal sourced products a safety risk, but also they simply are no longer necessary ingredients in manufacturing.

While I agree that eliminating animal-derived sources will improve the process, this claim is rather spurious.

Let's go through this one claim at a time:

Animal-derived products always carry a risk of contamination from adventitious agents, such as viruses and prions.

This is certainly true... there is always a risk, albeit small.  Peptones and other animal-derived media components are heat-inactivated.  Automation engineers code recipes into the PLC or DCS to make certain this happens.  After the animal-derived media components are added during cell culture media preparation, the entire media is virally-inactivated with a process similar to pasteurization.  Whatever risks were there are significantly diminished before the bioreactor is even batched.

In the case of antibody manufacture, the Protein A chromatography is the step that binds the antibody and throws everything else away.  To present the risks of viruses or prions in the final product as a risk - given the viral barriers and purification capabilities of modern biotechnology - is extremely irresponsible.

What about this claim?

Not only are animal sourced products a safety risk, but also they simply are no longer necessary ingredients in manufacturing.

Is it true that we fully understand the components of animal-derived ingredients?  I'm not aware that this is true.  Perennially, miniferm experiments confirm that cultures that use animal-derived sources produce higher titers than "veggie" media or fully-defined media.

There's that je ne sai quoi in animal-derived media that the cells seem to like.  Some think it is the shear-reduction capabilities; others think it is some magic ingredient that we have yet to fully characterize, but these process R&D managers aren't stupid.  If your process titers are low, and using animal-derived media can increase titers so that you can commercially manufacture the API... you use animal-derived media.  It's that simple.

All this said, The Dish is correct about eliminating animal-derived media components... but for the wrong reasons.

Elimination of undefined media sources will reduce process variability because you've eliminated a potentially huge source of variability.  I've seen titers swing 300% from year to year with no change in the manufacturing formula, no change in manufacturing execution, no change in anything but the lots of ingredients that we get. Using defined media will simply eliminate the lingering question of, "Is it the peptone lot?"

On top of that, variability reduction increases process capability.  Even if using fully-defined media loses out 25% yield to the best performing undefined media, you gain from the predictability of your process.

Now, you can schedule campaigns with greater confidence.  You can manage the supply chain.  You can be assure that there will be fewer disturbances to your manufacturing process... you can keep less inventory... and you can ultimately save in costs.

Variability reduction is the reason you want to eliminate animal-derived media components... not risk of viral or prion contamination.

Process R&D managers that "get it" understand that optimizing titer must be balanced with optimizing process capability.

Wednesday, September 19, 2012

Single-Use Bioreactors (there's the Mobius by @EMD_Millipore)

The concept of throwing away your bioreactor after a single use is quite foreign to me. My professional career grew up on stainless steel bioreactors that sat majestically in these uber-clean processing space in these plants.

But the disposable bioreactor concept appears to be gaining traction.

The idea is this:

You sterilize a plastic bag and you fill it with media. You can mix the contents of the bag. You can control temperature. You have pH and dO2 sensors hooked into a control system that maintains pH and dO2. You have everything that a stainless steel bioreactor has to maintain cell culture. Except when you're done with the operation, you get to throw away the bioreactor, whip out a new sterile bag and repeat.

What's the Advantage?

The most obvious thing is that you don't need to clean-in-place (CIP) or steam-in-place (SIP) the bioreactor... these bags come sterile. This means that you don't have CIP and SIP validation and the associated quality costs.

This also means that you don't have the piping, the valves, the boilers, the steam traps... all the infrastructure that goes into maintaining CIP and SIP systems.

CIP and SIP consumables don't need to be purchased. WFI systems can be smaller. Your Waste Kill isn't taking on all that acid, alkali and rinse solutions.

Staffing overhead to maintain and troubleshoot these systems don't need to be hired. The PLC logic or DCS recipes don't need to be coded. Control loops don't need to be tuned, so your automation engineers aren't as busy.

What about contaminations? Everyone knows from miniferm experiments that tube welding has low associated contamination risks. Everyone know that your failure rates go up the larger your tank. Is it true that success rates are higher with these single-use bioreactors?

What Stays The Same?

You're still going to need to execute, so your production organization ought to stay the same. You're still going to need to troubleshoot the process, so your Manufacturing Sciences process support stays the same. You still need scheduling, you still need downstream recovery and purification. All that stays the same.

I Still Have A Few Questions...

  • Do Single-Use Bioreactors (SUBs) decrease my plant shutdown times?
  • What is the turn-around time (from dirty to prepped) on a SUB?
  • What commercial products are manufactured with SUBs?
  • At what scale are SUBs proven?
  • What is the Total Cost of Ownership of SUBs?
If you want to know more, Dan over at EMD Millipore has some answers.

Contact Dan For More Info

Tuesday, September 18, 2012

How to Use ZOOMS (for OSIsoft PI System)

This was where our OSI PI Search Engine was as of 2008

In a single textbox, you could type in any set of words... just like Google.

And after you typed in a few concepts like:

  • Reactor1
  • Temperature
  • Concentration
  • Volume
ZOOMS could figure out that Reactor1 was a unit while Temperature, Concentration and Volume were process parameters.

Not needed, but if you felt like typing in a time-window (in OSIsoft PI-ese), it would simply show you the trend that you meant to see.

Perfect for people who don't have PI ProcessBook installed... which is basically management, scheduling, QA, instrumentation, process engineering... you know, everyone.

Get The ZOOMS Brochure

Dang @Novartis, 9 #FDA 483s Since 2009!? You Should Get With @fdazilla

I was reading this FiercePharmaManufacturing's Article on Novartis' CEO under a Regulatory Cloud.

So I got to thinking... how many FDA 483s are these guys getting?

It turns out, Novartis - across all their businesses has gotten at least nine 483s since 2009.

As Eleanor Roosevelt once said, "Learn from the mistakes of others.  You can't live long enough to make them all yourself."

Fierce is doing a good job tracking Novartis' quality problems. I'm just letting you know how to get your hands on their 483s.

Wednesday, September 12, 2012

Double Block and Bleed - Contamination

When filling up a bioreactor with media (colloquially called, "batching"), the media is delivered through a dedicated pipe.  When the batching is complete, there needs to be a way to contain the bioreactor... which is why there is a "near-to" block valve.

And the excess media (not required per specification) is sent to drain via a bleed valve that is on the non-sterile side of the block valve.

single block and bleed
Figure 1: Single Block and Bleed

This is pretty standard for fermentors and bioreactors... in the 1980's.

After the media line has delivered the media, it is dirty because the nutrient-rich media has coated the pipes and will promote growth unless cleaned.

And if you have single-block-and-bleeds, your cleaning and rinse solutions are going to pound against that block valve.  Should it fail, you're going to have a contaminated bioreactor.

Modern plants... those with the luxury of building from scratch install two blocks and two bleeds, hence
double block and bleed
Figure 2: Double Block and Bleed

In this scenario, when the media is delivered, both block valves shut simultaneously and both bleed valves are opened simultaneously.  The excess media is sent to drain via the valves farther away from the bioreactor.

Subsequent manipulations to that line (e.g. post-use integrity tests, cleaning, sterilization) are isolated from the bioreactor by two block valves... prophylaxis for disturbances to the sterile envelope.

Read About A Successful Contamination Response

Wednesday, September 5, 2012

Cell Culture Database - Batch Information

You work in biopharma. Maybe you're a fermentation guru... or a cell culture hot shot. Whatever the case... This is your process.
We muggles don't have the luxury of waving our wands and having protein fold themselves mid-air. There's usually a container where the process happens. processes happen in a unit
A time-window (starttime to endtime) is when processes happen.

Operators execute process instructions; these procedures is how the process happens.
The execution of process instruction results in an output. The output of the process step is the product and constitutes the what.
Lastly, the process (step) is given a name describing who the batch is.
It stands to reason that the who, what, how, when, where of a batch is characterized by:
  • batchid
  • product
  • procedure
  • starttime - endtime
  • unit
and fully describe batch information for cell cultures and fermentation.

Organize Your Cell Culture Data

Tuesday, September 4, 2012

#FDA Inspector Just Showed Up... @fdazilla

It’s Friday. 8:30am. You’re looking forward to a long, nice weekend. Then, you get a call from one of your associates. An FDA inspector is waiting in the lobby. She’s unannounced, she’s experienced, she’s eager, and she’s ready to go.

Do you know the feeling?

buy FDA Form 483