Meeting Discussion Point 2: IVIVC

[mattodd]

Background to Meeting on May 24th 2016 (#386).

Question: Based on what is known of metabolic clearance in vitro and in vivo, can we say there is a correlation between the two (meaning we can continue to measure in vitro values and have confidence that they are representative of in vivo: an in vitro-in vivo correlation (IVIVC). Question: If not, which compounds should we use to acquire more in vitro or in vivo data?

Resources:

OSM Series 4 summary wiki Lab books: Labtrove, Labarchives (Ed Tse, Chase Smith). Compound Database: All OSM Compounds in Google Sheet, interactive view on Cheminfo, direct download of Series 4 SDF.

Comments/suggestions below are welcome, as well as (briefly please) during the webinar meeting and its follow-up (#394).

Relevant background:

Whether we see an IVIVC for this series has been discussed in an OSM meeting and was the subject of a previous issue (#351), to which people are directed. In short, all relevant metabolic data for Series 4 are on the wiki here. There are only three compounds for which we have both in vitro and in vivo data (below). David Shackleford (Monash MIPS) suggested we were seeing some correlation, but we could use more data on a small number of compounds to firm this up. But which should we evaluate?

Considerations: Target compounds with good and poor measured microsomal clearance, from both amide and ether classes in this series? Focus more on those compounds for which we already have the most data (since this helps build a more coherent body of data, helping eventual publication).

Candidate compounds (shown below): MMV670944 (an amide that is potent, unlike MMV639565, with OK half-life) MMV668958 (an amide with OK potency and lower half life) MMV669844 (potent ether, for which we have a snapshot of in vivo efficacy.)

[Canute201]

I think more data would be helpful. I'd start with iv & oral mouse PK (as mouse is presumably your efficacy model) with cassette dosing, MLM, mouse microsomal Fu & mouse hepatocyte data, solubility probably worth getting too. I think it would be helpful to select a range of mostly the best compounds, but a range of structural subtypes would be good.

[mattodd]

Thanks @Canute201 - did you mean these data should be obtained for the compounds above, i.e. around 6? How many data points constitute a comfortable trend? Agree on the importance of mouse data @PaulWillisMMV ?

[mattodd]

Roger Bonnert (formerly an Associate Director and Project Leader at AZ) emailed after attending the meeting (reproducing with permission): "Re in vitro in vivo correlation I would ask why one compound doesn't scale rather than just test more, eg it could be blood stability (amide) or other extra hepatic."

I'd reply that arguing from single data points makes me nervous. How about we include an amide in a further round of measurements to see if both compounds behave similarly, then investigate whether there is some amide-specific issue?

Roger replied just before I posted this to say "I agree you might want more than one point but there maybe more simple explanations you can identify in vitro - blood/plasma stability (amidases?), I cannot remember if you're doing mic(rosomal) or hep(atocyte) (MHT: bit of both, here) but you'll clearly miss secondary clearance mechanisms with mic. Then in vivo you could see other extra hepatic clearances like renal or bilary Cl, though I realise there's not much you can do to easily predict. You can however design a subsequent study to collect urine for example."

Thanks for all this, Roger. So the suggestion is that rather than acquiring more in vitro data that we acquire blood plasma stability for amides and ethers. Any comments?

[mattodd]

Charlie Mowbray, Head of Drug Discovery at DNDi, emailed after attending the meeting to say (posting with permission) "For Vd very helpful to have some measured data, but once you have a few benchmarks then likely you can predict changes in Vd which are linked to logD and pKa."

Thanks Charlie. So perhaps 2-3 datapoints, would that be enough? Obviously on compounds for which we have logD and pKa.

[mattodd]

Neil Norcross from The University of Dundee emailed after attending the meeting. Neil wrote (posting with permission): "Regarding clearance, you might want to consider having lead molecules screened in hepatocytes for further understanding of both Phase 1 and Phase 2 metabolism"

I wrote back: We have such data on 3 compounds here, MMV669848, MMV669844 and MMV670936. Do these data provide the insight you were thinking of? In your view, ought we to acquire hepatocyte data on other compounds for which we have microsomal data?

Neil replied just now: "Regarding hepatocyte data, I might suggest to get hep data for compounds that are going into PK/efficacy, not necessarily for every compound, as any information is good information to guide future work.

Perhaps getting hep data is more useful when microsomal clearance is low but in vivo clearance is high and there seems to be a disconnect between in vitro and in vivo clearance/exposure. Just a thought.

Something also to bear in mind for in vivo work is in vitro plasma protein binding (PPB). How much of your compound is freely available to distribute to diseased tissues?"

Thanks for all this, Neil. If anyone has immediate comments on this, please go ahead. In the follow-up meeting (#394) we should finalise a list of frontrunners for scale-up synthesis, and we can then acquire extra data of this kind if we wish.

[mattodd]

Marvin Meyers, Director of Medicinal Chemistry, Center for World Health and Medicine, Saint Louis University attended the meeting last week and commented therein (posting with permission): "I would suggest that you consider doing an IV cassette PK study where you profile up to 5 compounds simultaneously in the same animal." Marvin's on Github as @mmeyers8 (not currently working, but coming).

Thanks Marvin. I presume you mean that we obtain the in vivo data for compounds for which we already have in vitro data? We should include a broad range of structures? Rat or mouse?

[mattodd]

As per analysis over at #400, the suggested amide MMV668958 could be replaced with MMV670767 if that is seen as being a better target for resynthesis.

[mattodd]

We need to complete the IVIVC issue - we are still seeking clarity on whether in vitro data are predictive of in vivo clearance.

It was previously suggested that MMV670944, MMV670767 and MMV897709 (rac MMV669844) would be appropriate (see newly updated scheme below). We have new data for these compounds (#484, not yet added to wiki section) and we have stocks. While the recent potency data differs a little from previous values, the clearance rates are largely consistent: the two amides have different rates, and the potent ether remains of interest given its in vivo activity profile. There’s a range of logD values (2.7, 3.1, 3.4)

Any comments on these three candidates vs., for example, any of the other frontrunners (#400)?

Then we’d need to assess what to measure. For consistency with existing data we ought to get rat data. There was previous support (above) for mouse data, and the new in vitro data was mouse and human (to align with previous in vitro data). However, Paul Willis at MMV had strongly suggested rat heps because of the eventual use of rat as the PK species.

The problem of comparing rat with mouse led, I think, to @mmyers8 to suggest an IV cassette study of 5 compounds in the same animal.

Do we need more hepatocyte data as per Neil’s suggestion? MMV670246 had already been flagged as needing such data.

Charlie Mowbray suggested acquiring a few Volume of Distribution values.

Current plan: rat hepatocyte data are needed for MMV670246, MMV670944, MMV670767 and MMV897709, then we aim for rat PK data that aligns with the data already obtained for the “knowns” MMV670246, MMV639565 and MMV669784.

Comments please by those with better knowledge than I.

State of Play.cdxml.zip