Breaking Bad in South Florida (5)

This is a work of fiction. Names, characters, businesses, places, events and incidents are either the products of the author’s imagination or used in a fictitious manner. If you find any resemblance to actual persons, living or dead, or actual events, deadly or lively, or actual molecules, carbons or heteroatoms, it is purely coincidental.

Part 5

(Here is Part 1, Part 2, Part 3, Part 4, Part 6)

Maybe I should make a little detour here and tell you about what our company was doing when not dabbling in the street drug manufacture – it is quite important to the story.

The main purpose of the company research was to formulate cancer chemotherapy agents into injectable therapeutic nanoparticles. We developed polymers composed of a random peptide sequence with lipophilic aminoacid residues and connected to a long polyethylene glycol chain. These polymers are bio-compatible and when dissolved in water they form pseudo-solutions composed of small fairly uniform micelle nanoparticles. They are acting like surfactant – with a greasy core and a halo of water-loving PEG chains projecting outside. When you have some lipophilic, poorly water-soluble drug that you want to formulate for intravenous infusion, you would dissolve it in organic solvent and add it to the micelles: the drug is going to partition itself into the micelle greasy core and after evaporating or dialyzing away the solvent you got an injectable formulation of a drug that is “solubilized” by being taken into the polymer nanoparticles.

Quite a few research groups developed this kind of system based on a polymeric excipient. A formulation like this is not stable and quickly unravels in the presence of plasma proteins. Our claim to fame and fortune was to include a metal-coordinating region in the polymer chain and finish the drug formulation by adding a small amount of Fe(III) salt: the added iron salt acted as crosslinker, it tied together the previously loose chains by metal atom coordination and we obtained stabilized micelles with the drug encapsulated inside. The nanoparticles could survive in the circulation and release their cargo gradually thus improving the drug residence time. The strength of used iron chelator is pH dependent and becomes weaker at lower pH and there is plenty of lactic acid in tumor sites. Tumors also have a voracious appetite for iron (with so many transferrin receptors on the surface of cancer cells) and their vasculature tends to be more leaky for the nanoparticles so there is an opportunity for improved delivery of a drug to the tumor sites with such iron-stabilized micelle drug formulation: The micelles make it to the tumor site, they preferentially fall apart there and the cancer cells get higher and more prolonged exposure to the chemo agent than they would get with a less elaborate formulation.

The controversial results with our technology came when I proposed and helped to develop an improved version of the polymer with stronger iron-binding properties. With the new polymers we saw the tumor mass discolored with iron that was part of the formulation, the tumor mass became visible on MRI and there were dense particles present in tumor cell vacuoles visible on electron microscopy. The company claimed that we can now observe the stabilized nanoparticles as they were making their way inside the tumor cells and releasing their drug cargo inside the cell.

There were few problems with this claim: 1. We had no proof that the high-contrast particles seen on EM and MRI were the original nanoparticles making their way inside the cell. A more likely explanation would be that cancer cells gorge on iron present in the formulation and they can protect themselves from the overload by stashing away the excess iron in vacuoles in a form of iron oxide particles that are highly visible and ferromagnetic. We did not try to look closely enough to be able to distinguish what were these particles. 2. There was no direct proof that our nanoparticles could survive their journey in the bloodstream and make it to the tumor site in one piece, as claimed by the company. We also couldn’t tell a micelle loaded with a drug from a micelle where the drug had already leaked out while it circulates in the bloodstream, and again we did not try to look very hard. 3. In fact, we had no way of distinguishing the free drug in circulation from the drug still loaded in the nanoparticle since any method used to analyze blood samples invariably destroyed the nanoparticles. You would “observe” a better drug residency times that way if you are combining together the concentration of free drug plus drug resting inside the nanoparticles and you cannot tell them apart.

I did not pay much attention to the biology, formulation or PK data analysis initially – I was interested mostly in making the polymers in high quality, trying to fix the numerous manufacturing problems. I hoped that after ten years in business the research director knew what he was doing. (He did: his plan was to sell the company at the first opportunity to the unsuspecting buyers.) But my biologist friend convinced me to look closer at the presented data and the story about the technology that our management was making and I noticed there was quite a bit of hand-waving that connected a perfectly sound research on how the drug was formulated with the results we were getting from the test animals implanted with xenografts and then treated with our formulations. We were blind as to the fate of the loaded micelles and their exact release mechanism, and the management probably should not have been making some of the claims in the absence of hard direct evidence.

My biologist friend was with the company for over 8 years, and he was clearly the most driven and most creative biologist we had, and the most skeptic one too. He would always run the controls on his controls, to make sure he wasn’t fooling himself – he came to believe that the entire subfield of therapeutic nanoparticles is polluted by sloppy science and irreproducible results. He blamed not only the difficulty of analyzing the nanoparticles in vivo but said it also originated from the wishful thinking, the socio-economics of postdoc academic labor and the relentless grantsmanship getting in the way of doing good science. He thought too many research groups were cherry-picking data to support their favorite notion without really asking the hard questions and this was the main reason why the nanoparticle field stagnated and so few approved drugs came out of it.

This biologist was always working on the periphery, on a sort of “woulnd’t it be nice also” kind of side-projects that were underfunded and received little chemistry support at our company. I was puzzled why the management did not have him on the clinical candidate formulation instead, with all his inventiveness and drive. The research director’s explained to me that my friend was “not a teamplayer”.

One of the side-projects that this biologist was doing was targeting: the idea of decorating the nanoparticle surface with ligands binding to receptors present on the surface of cancer cells. As our micelle was quite elaborate, fragile and not easy to follow in vivo, he developed a stand-in for a micelle, a toy model, a form of PEG-coated quantum dots of a similar size that can be easily studied in vivo because they are stable and highly fluorescent (you can even follow them in the circulation within a live animal if you use near-IR emitting Qdots.) He worked out all kinds of interesting details about the targeting problem and wanted to publish it. He also convinced himself that the micelle story the company was pushing was cartoonish and that the nanoparticles should not get inside the tumor cell if they looked anything like what the company was claiming.

When his targeting research project got cancelled in the summer of 2015 and my friend was finally put to work on the formulation with which we were going into clinic, he used some of his new analysis techniques and quantum dot-derived insights to have a more detailed look at our micelles. What he found was not encouraging: The drug was leaking from the “stabilized” micelle quite readily (on the timescale of minutes) once injected into a mouse, and the micelle itself was aggregating with blood plasma proteins. His work with cell cultures also supported a conclusion that the drug leaked from the micelles first and only then got inside the tumor cell, and not as a part of the intact nanoparticle. He did a FRET pair study with our nanoparticles – again, the leak out was rapid in undiluted plasma.

When my biology friend presented his data, there was a great deal of consternation among the chemists. The data indicated that there might be serious trouble ahead with our clinical candidate, and maybe even with the entire technology platform… In the meeting room after he finished we immediately started discussing fluorescent probes that would help us to clarify whether the problem was real and if it was, whether it was manageable with what we had, and if it was conceivable that the leakage problem was specific to the particular drug with which we were going into clinic (the drug did have a hydrolytic stability issue, the hydrolyzed ring-opened form was more polar and already known to leak out of the nanoparticles). Our research director cut the discussion short with saying that he was not surprised and that he suspected as much based on the animal data, and that he will carefully consider what will be the next step.

This was on Friday noon. Our research director then set up one-on-one lunch meeting with our biologist for the next Monday. This lunch meeting did not happen because our biologist found himself downsized already by 9:30 am that Monday morning. He received two months of severance for his eight years with the company – and he just had a newborn son and his ailing wife was at home with the kid. The research director also told him that his study the on quantum dot targeting wasn’t going to be published even though the company wasn’t interested in it anymore.

This happened to him after our management had been pooping on his work for years already and cancelling his projects every time he got some promising results, and now they were firing him for the work they asked him to do, in order to suppress his data. And he recently had to endure more than eighteen months of having our CEO for a neighbour – the CEO was using a lab located right behind his office desk, stinking up the place and cooking drugs like a maniac. The biologist wasn’t really given any other choice but to go after the company.