The eryaspase stumble, or why controls matter

Drug delivery is one of those enabling technologies which - while by themselves not as glamorous as, say, a vaccine for COVID-19 - nonetheless can have a profound effect on the outcome of medications. At Fannin, we have from time to time dabbled in these technologies. A couple of years back, we had looked at using biomimetic nanoparticles for delivery of oligonucleotide therapeutics. That didn’t quite pan out. Now, in a reversal, we are investigating oligonucleotides themselves as the delivery guides. We’ll let you know if that works.?

Be that as it may, reports of new drug delivery technologies are always of interest. One of the more innovative ones that made the news this week is the use of cells to deliver drugs. In theory, the cells act as a depot for a drug inside the body from which it is slowly released, increasing the half life and reducing toxicity. A French biotech, Erytech, had been quite successful in encapsulating large protein enzymes inside RBCs using alternate hypotonicity and hypertonicity to blow up the cells, capture the drug through open pores, then shrinking the cells to trap the drug. Apparently, this gives the entrapped drug (after the loaded RBCs are injected into the body) a half life of a month - a quite remarkable length of time, and way better than most if not all drug loaded nanoparticles.

Unfortunately, this week it all came undone. Here is the tale of woe. And an interesting theory to consider.?

The background

Erytech had decided to use their technology to go after cancer. But they didn’t load the RBCs with standard chemotherapy drugs - they used an enzyme, L-asparaginase. Normal cells can synthesize the amino acid asparagine, making it non-essential in our diet. Some cancer cells cannot - making them dependent on a supply of this amino acid for survival. L-asparaginase breaks down asparagine, depriving the susceptible cancer cells of an essential nutrient, eventually leading them to apoptosis. A neat concept. And it seemed to work, especially in blood cancers like ALL. But Erytech aimed higher: they targeted metastatic pancreatic adenocarcinoma, literally synonymous with a death sentence, with an average life expectancy counted in months.?

Did it work?

Rather surprisingly, perhaps, it did at first. The Phase IIb clinical trial enrolled 141 patients, and treated them with standard of care chemotherapy, with 2/3rds of the patients in addition getting the loaded RBCs (called eryaspase). The trial was reported to be successful. In the 46 patients receiving standard chemotherapy alone, the overall survival (OS) period was 4.4 months. Addition of the eryaspase treatment to the chemo regimens of 95 patients increased this to 6 months, which is a statistically significant increase. Understandably, Erytech launched a Phase III trial, with perhaps a song on their lips and a spring in their steps, confident of a successful outcome and a regulatory approval.

But..?

This Monday (25th Oct), Erytech announced that in the Phase III trial with 512 patients, the difference in survival between the two groups (with and without eryaspase) was not found to be statistically significant. Adding eryaspase improved survival, but not enough to garner an approval. What followed was predictable: the company’s stock dropped steeply, and industry reports indicated that the company will change focus back to the tried and tested blood cancers.

So, bottom line, the cell delivery technique does not work, right?

Well, maybe. But there’s something interesting in the limited amount of data released by the company.?

(Most of what follows is mere speculation and wild surmises from small pieces of data - and should be taken with a large dose of salt).?

Let’s take a look at the data from the control arms for the two trials - unsullied by any addition of eryaspase. The reported median OS of the control arm patients from the Phase II trial was 4.4 months, but in the Phase III trial, it was 6.7 months for chemotherapy alone - a >50% increase! That’s a little surprising, even allowing for some difference due to random chance.?

(The patient inclusion criteria in the two trials are almost equivalent: adults with confirmed advanced pancreatic adenocarcinoma,who have failed one previous first line chemotherapy regimen. Although there are some small differences in the way the inclusion criteria are reported in the clinical trials database, that does not seem to be sufficient to result in this difference.)?

Then why did the patients in the Phase III trial do better?

One reason may be the chemotherapy regimen.?

In the Phase II trial, the standard of care was defined as either gemcitabine monotherapy or FOLFOX (5-fluoro-uracil + leucovorin + oxaliplatin).

In the Phase III trial, it became either gemcitabine plus abraxane or 5-FU + leucovorin + irinotecan.?

We can hypothesize that the gemcitabine combination therapy increased the OS (compared to the monotherapy). In effect, the patients did better with 2 drugs. This makes it challenging for a third drug (eryaspase) to improve the OS even more.

This is somewhat supported by the company’s press release, which mentioned that for patients who received gemcitabine and abraxane, there was no benefit from adding eryaspase.?

However, in the other group, addition of eryaspase improved the OS from 5.7 months to 8 months, which is closer to what was seen in the Phase II data.?

So, eryaspase is useful, but when added to a poorer chemotherapy regimen?

That’s what a superficial look at the data suggests. In effect, as standard chemotherapy became better, the benefit provided by eryaspase went away. As the company delves into the data more, they may, of course, come up with a better hypothesis.?

But why did the standard of care change between the trials?

That’s an interesting question. Gemcitabine alone had been the standard of care for first-line treatment for advanced pancreatic cancer for more than two decades. However, the MPACT trial published in 2013 showed that gemcitabine-abraxane offered significant survival benefits over gemcitabine alone (albeit as first line therapy, not second line, but the impact on patient management was the same). Erytech's Phase IIb trial was started in 2014, and understandably used gemcitabine monotherapy as one of the options. By the time the Phase III trial rolled around in 2018, the combination therapy might have become more established. Eryaspase now had a higher hurdle to overcome: to show a benefit when added to a better performing chemotherapy regimen. Maybe that doomed it from the start. Maybe, if the Phase III had been performed a few years earlier, it might have shown a benefit over gemcitabine monotherapy in the Phase III trial and bagged an approval.

As they say, timing is everything.?

So, the bottom line is..?

It’s probably not a good idea to change the control regimen between trials. If the company had been aware of the change in regimen and moved swiftly to incorporate the new standard in the Phase II trial, they might have escaped with just a failed Phase II. But hindsight is 20/20, and maybe that change was not feasible.?

I still think the L-asparaginase angle is interesting. Eryaspase can probably find other indications where it improves outcomes, even in solid tumors. If it is, in fact, most effective when current chemo regimens are sub-optimal (e.g. in other difficult to treat cancers or as third or fourth line options) it may even succeed in demonstrating effectiveness there ... as long as the goalposts don’t move between trials.??