Answer ( 1 )

  1. It’s an exciting field! Sometimes you don’t want to know how the sausage is made. However, this is totally fascinating to know about.

    My adviser is very fond of saying that drug production is modern day alchemy. That is it’s the process of taking cheap organic matter and converting it into a product that is worth way more than gold.

    Chemical engineering is largely about understanding how one can control and manipulate processes to obtain a desired product. As with all chemical processes, there are additive, reactive, and removal steps that are selected.

    Drugs aren’t any different. I will warn you that I only know about the production process of biopharmaceuticals and I can’t go into details about small molecule productions. Biological drugs are particularly tricky since they are going into patients and there are insanely high standards on creating a reproducible, consistent product that is virtually pure, stable, and scaleable.

    In the case of biologics, we can largely divide the process into three stages:

    Production (fermentation or cell culture)
    Purification (separations)
    Packaging (fill and finish)
    Depending on the drug process and demand, the rate limiting step easily differs. 15 years ago, cell culture was the limiting step. Now that the yield has increased 1,000x for antibodies, purification became the limiting factor. When enough purification columns were finally built, they overwhelmed the fill and finish stages. The takeaway message is that all of these steps are important but some are more important than others depending on the process.

    Also important to recognize is that there are hundreds of quality control checks throughout the process. All of this occurs in a Current Good Manufacturing Practices (cGMP) environment. Anyone who has worked in manufacturing can tell you that there are numerous Certificates of Analysis (CoA) and Standard Operating Procedures (SOPs) that must be followed for Quality Assurance and throughout the process, samples are collected and analyzed to make sure that the process is in check. These documents are then examined by the FDA as part of the approval process and later inspections. As this is going into people and affects lives, it’s taken quite seriously. Variations away from cGMP may result in recalls.

    At this point, you can look at the figure below, give me an upvote, and call it a day.

    However, I’m only getting started. We’re talking about curing lives here.

    Production

    Cell culture is the goal of using recombinant technology to make cells convert one product into another. In most cases, this is either a feedstock like glucose or aminoacids like glutamine. By the point it gets to a manufacturing stage, the desired cell and desired product have been selected. Depending on the drug, different host organisms would be used but the most common are E. coli, Yeast, and CHO Cells.

    The goal of cell-culture or fermentation is to produce the most active drug / liter of culture-time consistently.

    The industry amusingly uses farming terms and those will be highlighted along the way.

    Unlike chemical synthesis where the ingredients tend to be consistent, cells may change or evolve during the course of the process. To prevent cell drift and to be ensure consistency, a single cell is grown in a small batch and then is saved into hundreds of culture vials and frozen away as a MASTER CELL BANK, with the assumption that they are all going to be the same. From this master lot, a working lot is then made using a similar process.

    When a batch is started, one of these vials (containing 2e6 cells) is thawed and seeded into the media. Once growing, the goal is to grow the cells to the highest density at the largest volume in the shortest amount of time. The best strategy is to keep cells growing at the log phase for as long as possible and hit the stationary phase in the largest reactor. Start too small and you fall into the lag phase; go too high and you have to to passage more frequently.

    The best way to do this is to constantly cultivate cells as they reach a critical density. In the manufacturing campaign I worked in, we passaged every 3 days using a 1:5 split from 200 mL -> 1 L -> 5 L -> 25 -> 100 L -> 500 L. The entire process took us 2 weeks before going into the 500 L. Larger companies like Merck and Pfizer have 10,000 L reactors. The timescales differ depending on the host organism.

    This is a “small” 3 liter biostat. Costs around $10,000 on ebay.

    A medium sized E. coli reactor

    A big one. (It goes underground)

    During the growth process, it is incredibly important that the cells are in an environment where they are willing to produce your drug. They are like any manufacturer so they need food, an appropriated temperature environment, and air.

    Several conditions are often measured and they include

    pH
    Oxygen intake and levels
    Glucose intake and levels
    Amino acid intake and levels
    Cell Density
    Temperature
    Pressure
    Stir-rate
    Like humans, if your neighbors are sick, you might get sick too so if any of these process measures fall out of range in the defined SOPs, the entire culture might begin to die at once and your drug gets destroyed. People monitor this very very carefully. To counteract imbalances, fed-batch approaches are often used by adding in acids or bases, air/oxygen, and other feeds.

    Equally important is the source of the media. The media used is a carefully designed buffered solution that includes necessary salts and chemicals that maintains the appropriate osmotic pressure and pH. They also have to come from FDA approved chemicals which are free from animal-derived products that may infect humans.

    When the cells reach their maximum density and achieve a stationary phase, the cells are then it is time for HARVEST!

    The all important drugs gets stored in cells via different mechanisms. Typically, E. coli cells store the protein internally. Mammalian cells typically excrete their products. Thus, the harvesting protocol differs.

    As E. coli cells store their drugs, we’re interested in harvesting the cells themselves. Using a giant centrifuge, the cells are pelleted down. To release the proteins, the cells are feed into a high pressure chamber to burst open the cells to homogenize the product. This lysate is then passed on to the purification team.

    Mammalian cells are different and in this case, we want the supernatent. We can clarify this supernatent by either centrifuging the cells and collecting the supernatent or using depth filters to trap the cells.

    Purification

    Purification is the process of taking this mixture of proteins, sugars, lipids, and molecules, and isolating and cleaning the desired drug product.

    The goal is to collect the most product with highest purity per column consistently.

    At our disposal are a variety of separation techniques which mainly use Liquid Chromatography. Many of you should be familiar with your highschool chemistry lab which separated colors from a leaf using a piece of paper and solvent.

    This is pretty much the same thing except 1,000,000x harder. Also people die if you mess up.

    The essential idea of separations and purification is to use certain physical and chemical properties of your mixture to either retain or remove components. These properties include:

    Size
    Charge
    Hydrophobicity
    Affinity to certain compounds
    We did an initial step by removing cells and cell guts using centrifugation earlier. Most pharmaceutical drugs will use an Affinity Chromatography (AC) step to isolate the drug. In antibody purification, Protein A will bind to the Fc region of the antibody and retain the desired product.

    Size Exclusion Chromatography (SEC) or Gel Permeation Chromatography separates out proteins by size. Small molecules will be trapped in the pockets of porous particles while the larger molecules will run through the spaces.

    Ion Exchange Chromatography (IEX) uses charged beads trapped in agarose or cellulose which holds on to ions. As a protein of certain charge flows through, the protein binds to the column. Then, using changing salt or pH, the interactions between the column and the desired protein weakens. The typical resins in order of decreasing pI are Q (quaternary), DEAE(Diethylethanolamine), CM (carboxymethyl), and SP (sulphonyl).

    Hydrophobic Interaction Chromatography (HIC) takes advantage of the hydrophobicity of proteins. At high salt, these interactions are particularly strong and using a decreasing salt gradient, the less hydrophobic proteins fall off.

    Like before, here are some representative images of the various sizes involved with purification.

    This is one we use in the lab. A cheap AKTA goes for $16,000 online.

    Here is a medium sized one. The frame costs around $8,000

    Here is a big one

    The limiting factor in purification is the resin. An inexpensive affinity resin like Ni-NTA will cost around $1,000 / Liter. Protein A or Strep-Tactin costs >$10,000 / Liter. As a result, you are limited by the maximum capacity of the column and the yield of your protein after the individual cleaning steps.

    Unlike fermentation and cell culture which is driven by cost per run, purification is a fixed capital expense. That is, 1 run vs. 10 runs will essentially cost the same (sans operator costs) since the expensive components are the rig and the resin. Unlike the production stage, each run is a fixed amount of time; the typical purification scheme will take a week.

    For validation purposes, the FDA will mainly be concerned about the purity and consistency as they want manufacturers to show that they are aware of every component that will go into the patient. As the individual clinical trials go by, the expectation is that the purity and reproducibility continues to improve.

    To illustrate how stringent we approach this, below is an example of a “shoulder” which shows a slight contamination of a dimer. This would be considered a red flag in the eyes of the FDA.

    Packaging

    The purified protein is the drug but what Pharmaceutical companies sell and the FDA approves is the drug product. For small molecules, this will include the pill and all of its excipients. For biologics, it’s the vial. Decisions in this process determines the drug shelf-life and delivery. This is the process that I know relatively little about so I am likely missing entire processes in this section.

    The goal of fill and finish is to minimize defects in creating the drug product.

    Typically, this follows the process below of sterilizing vials, filling the vials, lyophilizing the drug, and finishing the capping and labeling process.

    Fill refers to the processes of adding the drug into the vial. An essential portion of the step is the formulation of the drug ie. what salts and buffers are used to stabilize the drug. This influences the shelf life and if there is any pain from injections. It may also influence whether or not there are any losses from binding to the glass vials.

    Matt Harbowy has a wonderful answer on how this process is done precisely What type of machine could I use to precisely manufacture pills so that they contain exactly 1 milligram of active substance in each pill?

    The mixed protein product is dispensed into the vials using a machine as shown.

    In certain cases, the liquid mixture is capped immediately and stored at 4C. However, for drugs that need to be shipped overseas or for longer uses, we can freeze-dry the mixture by lyophilization. The vials are placed under negative pressure and frozen to cause the water to sublime without going through a liquid-gas transition which may damage the drugs.

    Finish refers to processes of adding the caps, sealing the vial, labeling, and inspection.

    Here is one representative photo of “capping

    Getting Approval from the FDA

    The OP asked to avoid taking about approval but the manufacturing process is essential to the approval process. A large part why natural drug products aren’t touched by drug companies and approved by the FDA is because the processes to reproducibly synthesize pure product is incredibly difficult.

    All of these processes are done in a clean room environment similar to Silicon manufacturing and they go from dirty to cleaner environments. The cell culture environment that I worked in was a Class 10,000 (ISO 7) and by the time it reaches purification and fill and finish, it will drop to Class 1,000. Contamination checks for mold and endotoxins are frequently done.

    I heavily emphasized consistency. When a drug goes up for approval, the FDA will ask for a demonstration of 2 beginning to end runs at the full scale. For this to happen, construction on a full scale plant and process will start 2 years before regulatory filling which means a company will typically be paying for its plant in the middle of its pivotal phase III trial with no idea whether or not the drug is going to work and ultimately approved. This is a relatively small cost compared to the actual running of the trial but it’s a cost that often gets forgotten when we talk about drug development.

    So that’s that. Drug manufacturing is the process of making small and “cheap” chemicals into an expensive specialty chemical that will go into a human being. Through a variety of steps, mainly, production, purification, and packaging, our goal is to do this reproducibly without defects with the knowledge that errors costs lives.

    It’s a constantly evolving field as people try to find ways how to make these drugs more inexpensively without sacrificing the quality that is required for these drugs to work.

Leave an answer