Friday, March 29, 2013

Did We Find the Holy Grail for Cancer?

A year ago, the American Association for the Advancement of Science (AAAS), through their website:, is reported the of a discovery of a drug that can kill all tumors.

It breaks down like this:

There's a protein called CD47 that tells your immune system, "Don't eat me."

Our immune system patrols our bodies looking to find and destroy foreign biological entities.  And apparently, one of the ways that our immune system recognizes foreign vs. self is this CD47 protein.

The discoverer of CD47, Stanford biologist Irving Weissmann, found that normal cells express CD47, but that leukemia cells express CD47 in higher concentrations.

Some blood cancers have been cured with anti-CD47... an antibody that binds to CD47 thereby blocking this "Don't eat me" signal emitted by cancer cells.

But that's not the news.  According to Weissmann,

What we've shown is that CD47 isn't just important on leukemias and lymphomas. It's on every single human primary tumor that we tested.
You can read about the simple A/B experiments they did with anti-CD47, but the basic gist is that they've shown in vitro that anti-CD47 helps macrophages recognize cancer cells.

They have been able to cure cancer of all types from the feet of mice.

The only question remaining is whether or not this works in humans...

Friday, March 15, 2013

Antibodies (mAbs) were 5 of top 8 best selling biologics

According to @CellCultureDish, 8 of 20 top selling drugs from last year were biologics.

#1. Humira (adalimumab)
#2. Remicade (infliximab)
#3. Enbrel (etanercept)
#4. ~not biologic~
#5. Rituxan (rituximab)
#6. Lantus (insulin glargine)
#7. Herceptin (trastuzumab)
#8.~not biologic~
#9. Avastin (bevacizumab)
#10.~not biologic~
#11.~not biologic~
#12. Neulasta (pegfilgrastim)

Source: Genetic Engineering and Biotechnology News

It's also interesting to point out that 5 of the 8 are monoclonal antibodies (mAbs). You can easily tell from the chemical name of the biologic because the chemical name ends with -mab.

Anatomy of the Antibody

For the non-biogeeks out there, monoclonal antibodies are Y-shaped molecules that are naturally produced by our immune systems to fight off foreign germs. Picture doing the "Y.M.C.A" dance and you're doing the "Y".

YMCA village people
The Village People emulating the shape of a monoclonal antibody

The significance of antibodies is that they are highly specific, meaning that they will bind to one target and only that target; it's possible because the antibody's molecular "hands" fit only one "glove" (specific antigen).

So the reason why monoclonal antibodies are so useful in medicine is because they can hit the targets that you want and nothing else... so-called, "Smart-bombs." And you'll see a lot of mAbs in cancer treatment (since you want to target only the cancerous cells, but not the healthy cells).

Equivalent of when an antibody comes across an antigen it doesn't recognize

mAbs can be divided into two regions: Fab and Fc.

Fab stands for fragment (antigen-binding). The part of the antibody that binds to the antigen. The Fab is basically everything from your shoulders up.

Fc stands for fragment constant: the rest of your body from your shoulders down to your legs. It's not really constant as it can be one of several classes in humans; but relatively, it's constant.

Early antibodies were engineered from mice. In very simplistic terms, you introduce a foreign biological molecule (antigen) to a mouse. The mouse's immune system will naturally fight off this foreign molecule by producing antibodies, and voila, you have yourself a mAb that can target your antigen. This mAb is 100% murine.

The problem with murine antibodies as medicine is that the human immune system recognizes it as foreign and will reject it. So to be useful, the antibody needs to be "humanized."

Humanizing an antibody means replacing mouse-Fc region with a human Fc region and then trying to make as much of the Fab region as human as possible.

mAb Nomenclature

There's actually a nomenclature for monoclonal antibodies that can tell you from the name, how much of the antibody is murine and how much is human:

diagram mab antibody

In the above diagram, the light-blue antibody is from a 100% murine (-omab). The darkish-red antibody is 100% human (-umab).

The genetically engineered antibodies are the ones that are mixed in color. A chimeric antibody (-ximab) is one where from the "elbow" down is human, and from the "elbow" to the fingertips is murine. Drugs like Remicade (infliximab) and Rituxan (rituximab) are chimeric.

A humanized-antibody (-zumab) is one where only the "fingers" are murine and the rest is human. Drugs like Herceptin (trastuzumab) and Avastin (bevacizumab) are humanized antibodies.

While the more humanized an antibody is, the less rejection it gets as a viable drug, there appears to be no obvious correlation from a revenue perspective (see list above).

What's this have to do with cell culture? Antibodies are secreted by genetically-engineered cells that are grown in cell cultures.

Related posts:

Follow me @zymergi

Tuesday, March 5, 2013

The Crux of Biologics Manufacturing

Whether you are making a biologic or a biosimilar, the manufacturing process (for the most part) is the same.

You're going to have cell culture or fermentation in either as a batch or continuous process.

You're going to have a recovery/purification train of 3 or more steps where you're purifying the drug and getting rid of unwanted biochemicals.

You're going to freeze that into huge ice cube and ship it to your fill/finish facility where they're going to thaw it and dispense the drug product into vials.

What biologic you make and how hard it is to make was decided by your Process Development group years ago when they were poking the DNA into cells, growing them and picking the "best" one. (This is a big-time simplification, the wonkish version of expression system selection, cell line development, cell line engineering, platform development, clone selection, process characterization/validation... etc.)

What decides whether you're making bevacizumab or rituximab?

If you inoculated a bioreactor with the cells that were transfected with the anti-VEGF gene, you're going to get bevacizumab.

If you inoculated a bioreactor with the cells that were transfected with the anti-CD20 gene, you're going to get rituximab.

It goes without saying that you must follow the recipe/manufacturing formula for the respective cell lines, which will contain minor differences across upstream and downstream.

The crux of biologics manufacturing is cell line development and what makes biologics manufacturing hard was doing the cell line development right.

What decides whether you're making Rituxan® or Reditux®?

If you inoculated your bioreactor with cells that were transfected with the anti-CD20 gene by an IDEC scientist, you're going to get the biologic: Rituxan®.

If you inoculated your bioreactor with cells that were transfected with the anti-CD20 gene by a Dr. Reddy's scientist, you're going to get the biosimilar: Reditux®.

What no one knew a decade ago (2003) was whether or not the Dr. Reddy's cell line could produce a molecule identical to the IDEC molecule. And even if they could, would this molecule be as safe and as effective as the FDA-approved IDEC molecule?

Is Making Biologics Hard?

This is a very broad question. Is putting cells into a bioreactor and watching them grow hard? ...because that's how the active pharmaceutical ingredients get made.

Is pouring the HCCF down chrom columns and hard? Broadly speaking, I'd say there's no shortage of people who can execute a large-scale biologics manufacturing process.

Is cell line engineering, platform development, clone selection tough...etc hard?  The grunt work can be done by entry-level scientists/engineers, but you probably want Ph.Ds leading the team to ask the right questions and to cast aside the technical road blocks.

Can Amgen take market share from Roche, Lilly, Abbvie and Janssen? Sure. Amgen has the right people, the resources and leadership.

Is making biologics so difficult to make that no one else can do it?  I don't think so. There's already generic versions of Aranesp, Neulasta and Neupogen in ex-US markets.