[procarch2019]

I work in the OT/automation space for pharma. Cell and gene therapy is crazy amazing.

One thing I didn’t really see mentioned here is length of treatment. They have to collect patients sample at hospital or other facility, transport to production facility, grow sample, inject sample with vector, grow some more and then reintroduce. The real sci-fi moment will be when they get to the point they can treat in a day.

I think most are 3+ weeks sample to treatment right now. Still amazing though.

[puzzlingcaptcha]

The 'vein-to-vein' time (apheresis to infusion) is being shortened all the time, and you can already manufacture CAR-T cells in 24-48hrs, but it still takes about a week to clear quality assurance and release the product (microbiology, integrated copy number, integration site testing, autonomous growth potential etc). The reason being mostly that you don't want to accidentally replace patient's B cell cancer with a T cell cancer of your own making.

[procarch2019]

Hah, you don’t have to tell me about validation and quality. Yes, they usually account for a significant time of anything in the pharma world.

[derefr]

> The real sci-fi moment will be when they get to the point they can treat in a day.

How do you imagine that would work?

Some closed-pipeline machine that lives in the hospital and automates sample → modified sample culture cycle?

Or something stranger, e.g. some kind of injectable (maybe prokaryotic?) cells that actively swim around looking for L-lymphocytes to vectorize through bacterial horizontal gene transfer — such that the whole process happens in vivo?

[puzzlingcaptcha]

> Some closed-pipeline machine that lives in the hospital and automates sample → modified sample culture cycle?

That's already a thing, for example Lonza's Cocoon platform or Miltenyi's Prodigy (both are often used for on-site manufacturing).

[panabee]

would you mind sharing what you think are the largest bottlenecks to (1) 24-hr turnaround and (2) CAR-T therapy for < $10K?

[puzzlingcaptcha]

I think that's near impossible to achieve with an autologous product. To meet those two goals I think you'd need to have an allogenic product manufactured and QCd in advance in a large batch, which you could thaw and infuse on demand. The problem with that approach is histocompatibility (HLA) matching, but several companies are working on that.

[panabee]

thanks for sharing. this is such an impactful space. hopefully someone one day can lower the cost to $10K or ideally $1K.

[orochimaaru]

Considering leukemia treatment today is 2 years for girls and 3 years foe boys, 3weeks is several magnitudes worth of improvement.

[chromatin]

You're quoting the length of Course V Maintenance in the CALGB 10403 study (or similar pediatric COG studies on which that was based), which is misleading in that it is (a) applicable only to ALL (b) patients under 40 y (c) only the final of five courses of treatment.

That being said, you're directionally correct and certainly, shorter duration immunotherapies represent a sea change.

[puzzlingcaptcha]

The treatment itself is a one-day procedure (although the patients will be generally kept under observation for CRS for a few days and then regularly monitored for response), the 3 weeks is the time needed for manufacturing/QC. In that time patients receive bridging chemotherapy until the CAR-T product is ready and unfortunately some do not make it.

[jfarlow]

There are a number of companies working on 'in vivo' deliveries for CARs. Oftentimes using the same tools as proven out by the Moderna vaccine.

[puzzlingcaptcha]

Yes, but that way you lose control over the dose, and to an extent over CAR-T characteristics. CAR-T therapy is usually used in patients who already had multiple rounds of chemo and their immune cells are generally not in a great shape. Even with 'traditional' CARs you occasionally get manufacturing failures since the cells are too exhausted to expand in vitro or have already lost their effector functions.