Mitigating ADC Toxicities with Linker-Payload Design

Hello everyone and welcome to today’s webinar, Mitigating ADC Toxicsities with Linker De Risking Downstream ADC Development by Optimizing Preclinical Assays with PEG containing Linker Payloads.

I’m Cassie Saltman of LabRoots and I’ll be your moderator for today’s event.

Today’s educational web seminar is presented by LabRoots and brought to you by Vector Laboratories. To learn more visit www.vectorlabs.com.

We encourage you to participate today by submitting any questions you may have during the presentation.

To do so, simply type them into the ask a question box and click submit. We’ll answer as many questions as we have time for at the end of the presentation.

Also, if you experience any technical issues, you can utilize this q and a box, and we will assist you directly.

I’d like to now welcome our speaker, Matthew Giese, senior scientist with Vector Laboratories.

Matthew, you may now begin your presentation.

Okay. Thank you for the introduction.

Welcome, everyone. Thank you for attending. Hope you’re having a great day. My name is Matt, and this webinar will be about using strategic linker payloads designed to mitigate off target toxicities of ADCs.

The balance between efficacy and tox is a bit of a tug of war, and this webinar is gonna focus on tox because many of the potential underlying mechanisms are influenced by the physical chemical properties of the conjugate. And the linker payload can have a substantial impact on those properties. The chemistry group here at Vector makes PEG based cross linkers, and this webinar will present examples where ADC developers have used cross linkers such as these to optimize ADC performance in assays that can suggest downstream tox liabilities.

So first, we’re gonna start out with a recap of some ADC basics. We’re then gonna present some examples with linear PEG stretchers, and this will be followed by examples with orthogonal PEG modifiers. And then we’re gonna wrap it up with an overview of vectors, d PEG based building blocks, and cross linkers that you could use to synthesize conjugates such as these. K. So let’s go.

ADCs are indeed the most successful class of actively targeted drug delivery systems, and this table shows the clinically approved ADCs. There’s been a lot of success with passive targeting in drug delivery, and there’s been some success using other targeting vectors such as peptides, but using antibodies as the targeting vector has yielded the greatest number of clinically approved ADCs.

However, as this table reveals, treatment related adverse events are not uncommon, and they can result in dose conduct dose reduction or discontinuation in the clinic. A twenty twenty three meta analysis found that in the one hundred and sixty nine clinical trials conducted since two thousand, which involved close to twenty five thousand patients. The incidence of treatment related adverse events was ninety one percent for all grade and forty six percent for grade three or greater. Furthermore, the failure rate in the clinic is not insignificant.

And of the ninety seven trials since, terminated since two thousand, eighty four percent were in phase one or phase two. And for the vast majority of these, it was due to lack of efficacy at the maximum tolerated dose or due to safety and tox concerns. So while ADCs have seen some success, there’s substantial room for improvement, particularly with respect to tox and safety. Now the reasons for failure due to tox in the clinic are manifold, but this webinar is gonna focus on the underlying mechanisms of targeted drug delivery with antibodies, and particularly the the preclinical assays that can be used to suggest something is amiss.

So targeted drug delivery is often selective for a tissue and not specific. It’s very difficult to thread that needle and get the therapeutic to only one location while avoiding all others. So the goal of targeted drug delivery is to tip that balance and reduce a therapeutic systemic exposure while increasing its exposure at the site of action. However, the reality of targeted drug delivery with antibodies is a little bit different. And while tumor exposure is generally increased, antibodies are taken up and processed in other tissues. For instance, highly perfused tissues like the skin, GI tract, and muscle are very efficient at catabolizing antibodies.

In addition, organs with high levels of professional phagocytes like the liver, spleen, and marrow, they’re very good at processing antibodies and their associating immune complexes.

In addition, the PKPD of the free payload must be considered. It is coming off of that delivery vehicle eventually. So all of these factors can contribute to these somewhat common toxicities that are seen across the different ADC platforms. So let’s review the underlying mechanisms of ADC internalization and payload release.

So first, this is the typical MOA given for an ADC. You have the conjugate, the tumor, and its target on the tumor cell surface. The ADC is administered via IV. It extravasates from circulation, diffuses through the tumor, and reaches its target on the tumor cell surface.

After binding, it undergoes internalization and trafficking through the endolysosomal pathway. And during this process, the antibody backbone and cleavable trigger are catabolized. This results in release of either the linker payload or the free payload, and then they can escape the lysosome and engage their intracellular targets and kill the cell, in accordance with the cytotoxin’s mechanism of action. In the field of antibody therapeutics, this is what’s called target mediated drug disposition, and it’s one of the processes responsible for the clearance of the antibody.

In the field of ADCs, this is also a process that’s responsible for payload release from internalizing antibody drug conjugates. However, this is an oversimplified picture. Instability in circulation can result in premature payload release or linker payload deconjugation. This is not a huge deal if these entities diffuse into the tumor.

The tumor itself is not a vacuous box, but is packed with vasculature and lymphatic vessels and ECM components.

Fortunately for the ADC, the vasculature is aberrant and leaky, and this can facilitate extravasation.

However, the densely packed collagen fibers and, ECM proteins can make tumor penetration difficult.

During residency in the TME, conditions and enzymes can also result in extracellular payload release, and this may or may not be desired. The ADC can bind to the neonatal Fc receptor, and this can salvage it from the endolysosomal pathway and recycle it to the cell surface, preventing catabolism and payload release. But this depends on where that trigger and and linker are designed to release the payload along the endolysosomal pathway. Recycling and release are kinda competing mechanisms in the different cells.

The payload itself can also undergo efflux from the target itself, either by passive diffusion or by active PGP mediated transport, and this can decrease efficacy. Sometimes, because tumors are heterogeneous and they have cells that don’t express the targeted antigen. So now that extra cellular payload can diffuse into these antigen negative cells to increase efficacy via the bystander effect. But this is still an oversimplified process because pinocytosis and receptor mediated endocytosis are also both endogenous internalization mechanisms.

And these are antigen independent.

Pinocytosis, is a nonspecific fluid phase endocytosis, mostly in the endothelial cells lining the blood vessels. But, it’s very efficient at catabolizing IgGs and this can release the payload.

Receptor media to endocytosis, facilitated by FCGR binding can trigger internalization and payload release. And the payloads can kill the cells.

They can diffuse out. They can diffuse into antigen negative cells for a bystander effect. And because this is in the tumor, it’s all good. But now it’s fairly well established that less than one percent of the administered dose reaches the tumor and is processed in this way. So where’s the other ninety nine percent of it? Well, there are other tissues and organs, and the ADC and the payload can be carried into these other tissues and organs and subjected to the same mechanisms of internalization and processing.

Pinocytosis is very efficient in tissues rich in capillary beds with high blood flow.

FCGR mediated RME is hard at work in tissues with high levels of professional phagocytes.

In addition, the antigen is sometimes tumor associated and not tumor specific, and now it’s expressed at low levels in healthy tissues. So now the ADC can bind to that antigen in the healthy tissues, and it gets subjected to internalization and processing. And this can result in either payload release or it can be recycled to the cell surface.

So this is presumably the fate of the other ninety nine percent of that ADC. And regardless of where the ADC is taken up and processed, the payload has the potential to reenter circulation and enter other tissues until it is eventually eliminated. So the reason I think it’s important to review all this, particularly with respect to the vector linker portfolio, is that these other processes besides TMDD are influenced by the physical chemical properties of the linker payload. And those are influenced by the different properties of how you put it together and how you build your payload. So properties such as size, aggregation state, multimeric state, hydrophilicity, charge, and even spatial arrangement can affect things like extravasation, tumor penetration, cellular internalization, diffusion into and out of cells, interactions with cell membrane surfaces, and interactions with the various FCGRs.

And all of these processes have the effect, potential to affect Botox and efficacy depending on where and when they occur.

So the goal of ADC design is to get more of the tumor and less of it to the other organs.

And this is kind of a tug of war between TMDD and these other undesired non targeted processes. So I’m not sure this is exactly a zero sum game, but reducing the amount of the conjugate that is consumed by these other processes makes more available to the tumor and vice versa. So this webinar is gonna present examples where developers have used linker payload designs that incorporate a PEG to reduce these undesired processes or at least improve the performance of the ADC in an assay that can be impacted by these underlying processes.

So there will be a number of preclinical assays that’ll be covered in this webinar that could be used to predict downstream tox liabilities.

Excessive hydrophobicity may suggest that nonspecific uptake will be an issue.

Instability could suggest reduced delivery of the payload to the tumor or an increased systemic exposure of that free payload.

Antigen independent activity can indicate non targeted activity in healthy cells.

Accelerated clearance of the total ADC relative to the undrugged antibody suggests drugging the antibody has had a detrimental impact on its PK, usually an increase in nonspecific clearance.

Accelerated clearance of the intact ADC relative to the total ADC suggests either instability or accelerated clearance of the HIDAR species is a problem or a combination of both.

Tolerability is assessed by a loss of body weight greater than ten, twenty percent, such as tox. Less than ten percent or twenty percent might suggest a narrow therapeutic window.

It’s almost essential to run clinical pathology studies in discovery case.

Antigen independent activity can indicate nonspecific uptake by these on targeted tissues and, again, suggest tox. And biodistribution studies showing accumulation of the ADC or the free payload in non targeted tissues can suggest toxicity to those non targeted tissues. So this webinar will present examples where PET containing linker payloads have improved assay performance in one or more of these assays. And ideally, the more boxes that can be checked, the lower the risk of inter insurmountable tox issues later on in clinical development. I wish I could give you a guarantee, but I can’t. It’s just more about minimizing risk.

So there are a number of ADC features that can be optimized to improve efficacy and reduce tox.

This can range from the antibody itself to the cytotoxin to the conjugation chemistry to the release mechanism and incorporating a hydrophilic component in the Langer, in this case PEG, but there’s also PSAR or sulfamides is one of those important alterations.

Due to the combinatorial nature of an ADC, it’s often many changes that add up to a better performing conjugate. So while this webinar is gonna focus on the PEG component, it’s often done in combination with these other changes. And these examples will show the broad applicability of a PEG with different conjugation chemistries, different DARS, different payload classes, etcetera.

This color scheme will be used throughout, and the components in black, purple, and pink will be that’s kinda similar to the products that are sold in the vector linker portfolio.

So now we’ll present some examples where these linker payloads have improved performance in those assays, and we’ll start with some linear linker payload designs. These put the PEG stretcher between the antibody and the payload and are the original incarnation of PEG based linkers and antibody drug conjugates.

So this first slide is a compilation of studies on tallowine and tesserine.

Recall an overall accelerated clearance or a large delta between the PK parameters of the total antibody and the intact ADC can suggest something is amiss.

So tallowheme is shown on the left here, and this was site specifically conjugated to three different antibodies that get these three different ADCs. And these were evaluated in various preclinical and clinical studies. However, in all cases, there was either a large delta or a short half life and low exposure, and development of all of these ADCs was discontinued at some point. On the right is the redesigned linker payload. This is Tesserine, and it uses a different cytotoxin, s g thirty one ninety nine. It’s a self emollative spacer, and it uses m p peg eight for conjugation.

This was site specifically conjugated to two different antibodies to give roba t and kany t that were evaluated in various clinical trials. However, again, in both cases, there is either a large delta or a short half life and low exposure, and development of both of these was also discontinued at some point. In the studies on the left, it was found that conjugating tallowhene to the hemichysteines resulted in significant aggregation, and this could be ameliorated with that site specific conjugation strategy. However, this redesigned link or payload allowed hinged cysteine conjugation to be revisited.

So changing the conjugation site back to the hinged cysteines and targeting a different antigen provided this ADC, which now has a small delta, a long half life, and a good exposure. And this is Xiamonta that was approved in twenty twenty one. So this example demonstrates that this optimal linker payload design could be coupled with a conjugation strategy to improve the p k parameters that can sometimes be indicative of off target toxicities.

This is a good example from AstraZeneca showing that the inverse relationship between DAR and clearance can be mitigated with a linker payload that incorporates a PEG eight stretcher. In this case, the linker payload can be compared to tesserine, how it uses this less potent and less hydrophobic cytotoxin, SG thirty six fifty. The reduced potency can be compensated for by increased drug loading. And indeed, this linker payload design facilitated the construction of DAR two, four, and a ADCs with minimal aggregation.

In addition, PK studies in rats given equal payload doses found that the PK parameters were similar for all of the ADCs regardless of DAR, and they weren’t that much different from the unconjugated antibody. So this suggests that the high drug loading did not have a detrimental impact on the PK.

When these conjugates were screened at subcurative equal payload doses in xenograft models, they were also equally efficacious. So again, this indicates that the high drug loading did not accelerate that nonspecific clearance and decreased efficacy.

Finally, tox studies reveal that the maximum tolerated dose in monkey was five fold greater than the efficacious dose, and that’s compared to three fold greater for tesserine.

Furthermore, if you note the, difference in the PK parameters between the total antibody and the ADC, they suggest linker payload loss. However, the tolerability studies show that the penalty for this is not as severe with these lower potency payloads. So this example demonstrates that this optimal linker payload design with the PEG eight stretcher can reduce aggregation and accelerated clearance and improve the tolerability.

This example from Mavel shows that payload delivery and tolerability can be improved with a PEG four stretcher. And in this case, they use their proprietary disubstituted maliumid, which gives stable disulfide re bridged DAR four ADCs after ring opening. The comparator for this study is EV or PadSeq shown below. So it should come as no surprise when these two conjugates were evaluated for stability in monkey serum, EV demonstrated significant payload loss while Madwell’s ADC was much more stable.

The internalization studies were interesting. And in high expressing cells, both ADCs internalize similarly. However, in medium expressing cells, Madwell’s ADC internalized better than both ED and the unconjugated antibody, and that’s despite identical binding affinities. So that suggests that some property of the linker payload or conjugation, some property like the extra negative charge or the a ring or the PEDspacer affected either TMDD or antigen independent uptake.

And depending on which, this could suggest increased efficacy in tumors with low antigen expression or it could suggest increased TOX due to antigen independent uptake.

PK studies of both the ADC and the free payload indicated that despite the increased stability, MaBWell’s ADC was cleared faster and its plasma AUC was twenty percent lower than that of EV.

However, it also had a sixteen percent greater delivery of MMNEE E to the tumor. And I don’t show the data. It was also more efficacious in xenograft models. So this is a good example of that duality I was talking about. Up to now, the examples have suggested that increased stability means increased exposure means increased efficacy. But in this case, the opposite is actually true. And a reduced systemic exposure of the ADC appears to indicate greater efficacy.

So faster clearance and low exposure per se is not necessarily a bad thing depending on what is doing the clearing. And in this case, it appears that the linker payload has increased the clearance by the tumor, which is a good thing.

So this, ADC was also evaluated in monkeys, and the reduced systemic exposure compared to ED was also noted here, and it had a lower mortality rate. MadWell is currently in clinical trials with this candidate. And when compared to EV, it appears to have a better objective response rate and a lower occurrence of treatment related adverse events. So it’ll be interesting to see how this goes in phase three. Preclinical studies did show relatively high liver uptake, but, MapWell is using this linker design in their other clinical stage candidates. So this example demonstrates that this PEG four linker or PEG four containing linker can help improve payload delivery and tolerability and, probably increase the, therapeutic index.

This is a great example of ADC optimization from, AstraZeneca, and they show that stability and tox can be improved by a linker payload that incorporates a PEG-eight stretcher and this proprietary TOPO-one inhibitor. So all four of these, ADCs use the same proprietary TOPO-one inhibitor. And then for conjugation, they either use MC or MP peg eight, and the cleavable trigger is either GGFG or val ala. And then they mix and match the various components, and they look at them in these different set of assays. So the least hydrophobic, at least by HIC, is the second from the left.

The most stable conjugates are the two on the right. So link or payload deconjugation via retro Michael and stabilization via ring opening are kinda competing reactions. So you do get some link or payload loss with all of these. However, it’s minimized with the MP PEG eight. When evaluated for efficacy, the MP PEG eight GGFG is the least efficacious, although it’s difficult to differentiate the other two. However, tox studies revealed clear differentiation.

And this last example, this last combination with m p peg eight, val ala showed only mild and reversible hematology changes. So this last one is also AZD eighty two zero five, which is currently in phase one clinical trials. So this is a nice example showing that the different, ADC and linker payload components can be evaluated in these different assays to improve physicochemical properties and reduce tox in these preclinical assays that might suggest potential downstream tox issues.

This is a good example from cGen showing that stability, PK, and non targeted uptake can be improved with the linker payload that incorporates a peg four stretcher in this proprietary TOPO one inhibitor. So in this example, conjugates with this linker payload, LB, were compared to conjugates with this linker payload, DT.

The conjugates with LB had reduced HIC retention time relative to conjugates of DT, And they also exhibited a significant, reduction in liquor payload conjugation. Comes as no surprise.

When conjugates with this new linker payload were evaluated in PK studies, it was found that the concentration of total antibody was not that much different than the undrugged antibody. So that suggested this high drug loading did not have a detrimental effect on the p k parameters.

On the tox side of things, when conjugates were evaluated in an assay designed to predict nonspecific uptake by the liver, there was clear differentiation, and conjugates with LB were taken up seven to tenfold less than conjugates with DT.

This low nonspecific uptake was also correlated with favorable rat tox studies. And where rats were dosed with non targeting conjugates, they had hematology measures that were similar to vehicle, and there was minimal loss in body weight. So this example demonstrates that this linker payload design with the PEG four stretcher and the proprietary PUC TOPO one inhibitor can improve physical, chemical, and pharmacokinetic properties, and then it can also reduce non targeted uptake and myelosuppression, which might suggest reduced downstream tox liabilities.

This final example in this section shows that antigen independent activity of de crater antibody conjugates can be improved with linker payload designs that incorporate a pegging stretcher.

These examples come from different papers, so they degrade different POIs, but they both target HER2. They both use disulfide rebridging for DAR four ADCs, and they both recruit VHL. So So I think we can make some inferences with those caveats, of course. So on the left, both targeting and non targeting conjugates were prepared, and they were tested for POI degradation in cells expressing HER2. In the example on the top, it can be seen that there is POI degradation with the targeting conjugate. However, in the example on the bottom, there is also POI degradation with the non targeting control. So the authors suggest that this antigen independent activity is due to nonspecific uptake of that hydrophobic conjugate.

In the example on the right, PEG stretchers were incorporated into that linker payload to give a more hydrophilic linker payload design. And then this was tested for POI degradation in cells that both expressed or didn’t express the antigen. So in the example on the top, it can be seen that there is POI degradation in the HER2 expressing cells. And then on the bottom, there is no POI degradation in the cells that aren’t expressing HER2. So there’s no antigen independent activity. So this is a good example showing how these hydrophilic linker payload designs that incorporate PEGs can reduce this antigen independent activity that has a potential to give rise to off target toxicities.

So now we’ll switch gears and look at some linker payloads that incorporate these orthogonal PEG modifiers. And these put a PEG in an orthogonal shielding position so they can not only compensate for hydrophobicity, but they can provide shielding and maybe accommodate different functional groups if desired.

This is a good example from AstraZeneca showing that stability and efficacy can be improved by designing a linker payload and conjugation strategy to evade an offending systemic enzyme. Four different ADCs were prepared with this VALSIT trigger in different distances between the antibody and the trigger. In format a, it has no stretcher in the middle. Formats b six and b twenty four have PEG six or PEG twenty four stretchers. And format c, again, has no stretcher, but it places that PEG in that orthogonal shielding position. The Valsit trigger is susceptible to cleavage by circulating CES one c in mouse, and that was verified here for format c twenty four. However, CES one c has a very restricted active site access, and that can be leveraged by some of these conjugates.

In vivo stability studies found that formats b six and b twenty four with that long conjugation distance were extremely unstable. However, format a with the short conjugation distance was much more stable. But now that hydrophobic payload is exposed and it could present some downstream liabilities with development.

Format c, which also retains that short conjugation distance and places the PEG in the orthogonal shielding position, was just as stable as format a. So presumably, this increase in stability is gonna reduce toxicity, and it probably comes as no surprise that in xenograft models with circulating CES one c, efficacy also is tracked with stability. So while this is a species dependent phenomenon and not expected to be an issue in monkey or human, it does demonstrate this important concept of using linker design to leverage steric shielding from the antibody and hydrophilic shielding from the PEG somewhat independently, and this can prevent premature payload loss.

That same linker payload design was used in a former clinical candidate from OBI Pharma, and they demonstrated that FCGR affinity could be reduced with this linker payload. There’s been a couple studies that have shown that conjugating smaller linker payloads to the HIN cysteines doesn’t have a substantial effect on FCGR binding, but conjugating larger structures like pegs to that region can. And in this case, it was demonstrated that this linker payload did reduce FCGR affinity.

Given that these receptors are expressed on professional phagocytes, this has the potential to reduce uptake by or increase clearance from tissues containing those cells. So while it’s only speculation, this might be supported by the biodistribution studies. There was initial distribution to the highly perfused tissues, including the organs of the MPS with the professional phagocytes. However, Cmax was reached after only four hours. And after that, there was a rapid decline, and this was accompanied by a slow prolonged uptake in the tumor. And this resulted in tumor to tissue ratios greater than one by only eight to twenty four hours post dose. So while development of this conjugate was discontinued last year, it demonstrates this important concept of using the PEG to shield functional domains on the antibody that I might affect these off target uptake and contribute to downstream tox issues.

This is another clinical candidate from OBI Pharma, and they demonstrate that non targeted white blood cell interactions can be reduced. So they don’t disclose the nature of the, linker, but it could be similar to the previous example. When evaluated against dado DXD, it was found that it had a more safe, favorable safety profile in monkey, and it also had reduced interactions with white blood cells. And this can indicate a lower risk of myelosuppression later on in clinical trials.

Given that these, these white blood cells express all these FCGRs and the previous slide demonstrated that orthogonal PEG attenuated these FCGR affinities, this could be a similar linker payload design. So both the last example and this one demonstrate these types of studies that can be done to design linkers that minimize these non targeted interactions that might contribute to potential downstream tox liabilities.

This might be a good place to note that some developers like Zymeworks and Eli Lilly and Tubulus are using FC silenced antibodies in some of their clinical stage candidates, and this could be another way to attenuate that interaction.

We’re gonna spend a couple slides looking at some examples from Cgen since they do a really nice job of exploring how these linker variables affect the different assays. In this study, they show that clearance and non targeted uptake can be reduced with this orthogonal PEG modifier. So they made three different ADCs, In ADC six, it has the MC group for conjugation and an orthogonal PEG modifier. In ADC four, it has the MC group for conjugation. And in ADC five, it is the MC group and puts a linear PEG stretcher between MC and that linker payload.

When these ADCs were evaluated in PK studies, it was found that ADC four was cleared quite quickly, and this is not uncommon for DAR eight ADCs with this MMAE payload.

ADC five with the linear PEG stretcher was cleared even faster. However, ADC six with that orthogonal PEG modifier was cleared much slower, and in vivo efficacy also tracked with clearance.

Conjugating MC in cysteines is notorious for link or payload deconjugation, so that was explored as a possible contributing factor. However, evaluating these ADCs in, rat plasma revealed that they had identical rates of link or payload deconjugation. So that wasn’t that wasn’t, the difference here. However, when Sprague Dolly rats were administered a six MPK dose, IHC revealed significant differences in hepatic update. And the faster clearing ADC showed the darkest staining while the slowest clearing ADC showed little to no staining. So this example demonstrates that this orthogonal PEG modifier can reduce that accelerated nonspecific clearance and reduce nonspecific uptake by the kidneys by the liver. I’m sorry.

In a follow-up study, it was demonstrated that efficacy, exposure, and tolerability are affected by that length of that orthogonal PEG. In this study, they used the self stabilizing mDPR group to prevent linker payload loss, and then they varied the length of that orthogonal PEG. These ADCs all had nearly identical in vitro cytotoxicity, but they showed a clear differentiation in these xenograft models. And those with the PEG eight, twelve, and twenty four show the greatest amounts of TGI, and PEG eight and twelve appear to be slightly superior. This trend was reflected in the p k studies, and conjugates with the eight, twelve, and twenty four had p k profiles that were similar to vehicle.

This was also reflected in tolerability studies, and the ADCs would be eight, twelve, and twenty four were tolerated similar to vehicle with the PEG12 appearing to be slightly similar. So this is a nice example showing that orthogonal PEG modifier can be optimized to provide this best balance of efficacy, PK, and tolerability.

In another follow-up study digging into the mechanisms of what’s going on here, it was demonstrated that non targeted uptake by Kupffer cells, which are the resident macrophages in the liver, also depend on the length of that orthogonal PEG. In this study, non targeting conjugates were administered to the rats and the length of that orthogonal PEG was very, valid I’m sorry. They were it was an in vitro assay. It wasn’t administered to rats.

But it was found that this nonspecific uptake depended on the length of that orthogonal PEG. And ADCs with the PEG12 and eight were taken up to the same extent as the undrugged antibody, while those with no PEG or short PEG saw a two to threefold increase in nonspecific uptake. So ADCs with Fc engineered, regions were also used, so they didn’t have fcGR affinity and they behaved similarly. So this does not appear to be primarily a FCGR driven process, at least in kuffer cells.

But this study demonstrates that the orthogonal PEG can alter this underlying mechanism of antigen independent uptake, whatever it might be.

In a follow-up study, it was also found that the payload concentration in non targeted tissues depended on the length of that orthogonal peg. In this study, nontumor bearing rats were administered nontargeting conjugates, and the concentration of free MMAE in the liver, spleen, marrow, and plasma was determined by LC MS.

It was found that ADCs with the longer PEGs resulted in payload concentrations that had a nearly twofold reduction in Cmax in these non targeted tissues relative to the ADCs with the shorter pegs or no peg. All the ADCs also exhibited difference in Tmax, and it occurred at about six hours in the liver and twenty four to forty eight hours in the spleen, bone marrow, and plasma. So in this case, it was suggested by the authors that this rapid nonspecific uptake by the by the liver resulted in catabolism of the ADC and release of the free payload, which slowly diffused into these other tissues. So this study shows that this orthogonal PEG can be modified to, reduce this nonspecific uptake and catabolism of ID, ABC in this undesired, tissue, the liver. And then this can also decrease the concentration of the free payload in the off target tissues in the spleen bone marrow.

That same study demonstrated that the tolerability in clinical chemistry parameters depended on the length of that orthogonal PEG. And ADCs with the, no PEG or short PEG showed very low survival rates, while those with the eight, twelve, or twenty four, the animals showed, one hundred percent survival rates. It was also found that, hematology and serum chemistry depended on that PEG length. And while all all the ADCs showed some, neutropenia, those with the longer PEGs showed less myelosuppression and liver enzyme elevation.

So this is a nice, set of studies in these last, slides showing examples of how the orthogonal PEG can modulate these underlying mechanisms of nonspecific uptake that can contribute to dose limiting toxicities later on in clinical development.

This idea of using the orthogonal PEG to shield undesirable physical chemical attributes is also being adopted for other payload types. And in this example, it is a study from Tubulus showing that stability in PK of DAR8 exa ADCs can be improved with these orthogonal PEG modifiers.

So during the initial development of ADCs with TOPA one inhibitors, it was found that EXA could not be used as a payload because it resulted in aggregation at DAR eight. So DXD was developed, and this facilitated these construction of DAR eight ADCs. So that’s inherited, which is shown on the top here. This is the control in these experiments.

Tubulus was able to use their proprietary conjugation chemistry and these orthogonal PEGs that are capped with hydroxy dermis, and they were able to make these DAR eight exa ADCs with the valsit trigger in high yield and little aggregation. And this is TLP three until TLP five shown on the left there. When TLP five was compared to INHER2 in DK studies, it was found that total antibody concentrations were similar, but it should come as no surprise there was significant deconjugation with INHER2, but TLP5 remained quite stable. Evaluating the PEG lens, it was found that the PEG24 provided greater physical stability, and it also provided greater exposure.

So TLP five was compared to INHER2 in xenograft models, and it was also shown to be much more efficacious. Tubulus currently has two clinical candidates with this linker payload, and they’ve also published similar studies using MME as the overhead. So this example demonstrates that, that orthogonal PEG PEG modifier can be used to improve physicochemical and pharmacokinetic properties that might suggest, improved tox profiles later on in clinical development. And in this case, it appears that it’s the PEG twenty four hydroxy that is most optimal. In the Seagen case, it was the PEG twelve methoxy.

So to wrap up this section, well, there’s an example showing that the antigen independent activity of immunostimulant conjugates can be improved with these linker payload designs. And this is a study by Mersano.

In the conjugate on the left, a linear PEG stretcher is incorporated between the antibody and the payload, and it’s non cleavable. In the example on the middle, there’s an orthogonal PEG that’s capped with these very hydrophilic sugar moieties, and it has a cleavable dipeptide. And the example on the right, it has that same orthogonal PEG modifier with the sugar residues, and the cleavable dipeptide has been replaced with a cleavable ester and a PEG stretcher. And then these three linker payloads are conjugated to both targeting and non targeting antibodies to look for both antigen dependent and antigen independent activity.

So when the antigen, when the targeting conjugates were, tested for activity, they all displayed equal efficacy. So they did not differentiate on the level level of efficacy. However, when the non targeting controls were tested, the conjugate on the far left exhibited a con significant amount of target independent activity, and the conjugate on the far right exhibited the least. This was also reflected in the tolerability studies, and the conjugate on the far left showed the greatest amount of loss of body weight, while the conjugate on the far right showed the least. So presumably, this reduction in hydrophobicity of conjugate thirty seven relative to the other conjugates Minimize this antigen independent uptake by reducing this nonspecific uptake of the hydrophobic conjugate. So this is a good example showing how these hydrophilic orthogonal shielding pegs can minimize this antigen independent activity by minimizing nonspecific uptake of the conjugates.

This last one is XMT twenty fifty six, which is currently in clinical trials with a different, immunostimulant payload.

So we’ll wrap this up with some, examples from the vector linker portfolio that you could use to synthesize conjugates such as these or screen different conjugation strategies or linker variables if that’s desired. So most of this webinar focused on conjugation to hinge cysteines because there’s more precedent for the PEG, but site specific conjugation strategies are rapidly becoming the dominant approach, and they also incorporate PEG components. The example on the top left is Synifix’s technology for QuickChem at Modified Glycans. This is, ADC Therapeutics phase one candidate.

The example on top right is, Ambrex’s technology for oxime ligations at engineered residues. This linker payload is called Ambrastatin AS two sixty nine, and it is currently in three different clinical candidates. The example on the left is Sutro’s technology for ClickChem at engineered glycans, and this is a former phase one candidate of theirs. The example on the middle right is a structure disclosed in ProFound Bio’s patent application, and this is also for ClickChem.

And the examples on the bottom are for enzymatic ligations. The one on the left is from a preclinical study by Pfizer, and this is for transglutaminase mediated ligation. And the example in the middle is from Gene Quantum’s patent application for ISACS, and this one is for sortase mediated ligation. And the example on the right is from patents at, Talic’s patent application on antibody oligo conjugates.

And in this case, they use a short PEG to put DBCO on the oligo, and then they use longer amino PEG azides that are conjugated to the antibody via transglutaminase, and then they link those two via click chemistry.

So PEG components are certainly finding a place in site specific conjugation technologies. And oftentimes, it’s just to facilitate the conjugation, whether it’s to offset the hydrophobicity of DBCO or give the enzymes a little room more room to work. You do have to be able to make these things in the first place.

If you go to conferences or investor presentations, you’ll see that these hydrophilic components in the linker are increasingly regarded as trade secrets, and they’re not disclosed in these widely disseminated publications. But if you spend some time digging through associated patents, you can make some good guesses, and oftentimes, they do incorporate a peg.

These linker payload designs with PEGs are also becoming more common in patents with other payload types. For instance, with, metancinoids, TOPO one inhibitors, there’s, antibody oligo conjugates, immunostimulants.

PEG containing crosslinkers aren’t just for cytotoxins anymore, and Vector has a number of crosslinkers that you could use for whatever type of cytotoxin or payload you’re conjugating to your antibody.

Vector has two product lines. There’s building blocks and then there’s ready to use crosslinkers.

The Deepgram based building blocks are for you to build your own cross linkers with your multi step synthesis, and this is if you have some sort of proprietary reactive group or cleavable trigger that we don’t offer. And, yeah, you could just build it that way. We also have ready to use cross linkers, and these can help expedite the synthesis of your conjugate. And this is if you don’t wanna mess with the synthesis of the cross linker and rather focus resources elsewhere. For instance, the, SAR, the cytotoxin or developing the release mechanism or maybe developing the antibody itself.

This these can all be let’s see. These can also be particularly useful if you’re looking to screen different conjugation strategies or possibly screen different linker variables. And, again, you just don’t wanna mess with the chemistry of the different cross linker combinations.

So these are examples of vectors, building blocks that you can use to expedite the synthesis of your cross linkers with your proprietary reactive groups and property modifying functional groups. Now we didn’t cover chemistry in this presentation, but if you would like to go back and review these slides and, more importantly, the references and the associated supporting information, you can see how these building blocks were used to make the linker payloads in those examples.

These are examples of red vectors ready to use cross linkers if you don’t wanna mess with their synthesis. And these can allow you to expedite the synthesis of your conjugates.

So they’re available with a wide variety of reactive groups. One of them is typically an active ester for your amine containing payloads. And for PEGs, we’re generally looking at PEG4, eight, twelve, and twenty four, although others are available.

For conjugating to the antibodies, we have reactive groups for cysteine conjugations, oxime ligations, disulfide rebridging, and click chemistry.

The full color structures are catalog items available in high purity at kilo scale. The transparent structures have been made, but they’re not kept in inventory, so you’d have to contact the sales team for details. And, again, if you’d like to go back and review these slides, you could see the benefits of using similar types of linkers in some of those examples.

This matrix shows the different combinations of reactive groups available on header bifunctional cross linkers, and this is if you’re interested in something besides that active ester. The numbers in the slides or the numbers in this table denote the slides that use similar types of linkers in those examples if you’d like to go back and see it how they’re used. The structures denoted or coated purple are standard catalog items available in high purity at kilo scale. The products coated blue are not kept in inventory, so you can contact the sales team for details.

These are some examples of vectors ready to use sidewinder cross linkers, and these have that orthogonal PEG shielding structure. These are also available with, different reactive groups, but one of them is typically the active ester for mean containing payloads.

For the PEG lengths, we’re looking at PEG four for the stretcher component and twelve or twenty four for the modifier. And they’re typically built on a lysine or glutamic acid core, and the catalog products, the full color structures are available for cysteine conjugations while the transparent structures are currently in development. And, again, if you would like to go back and look at the slides denoted here, you can look at some of the benefits of using similar types of linkers in those examples.

This matrix shows the typical reactive group PEG terminate combos on some of these sidewinders, and this is with the reactive group as the TFPSter.

These numbers, again, they denote the slides. If you would like to go back and see the advantages of using some of these types of linkers in those conjugates.

The product coated purple is the only standard catalog item, but it is the most common maliumid active ester combination with the shielding MPEG. And the products coated blue, we have the building blocks around, but they’re just made to order, so you would need to contact the sales team.

Finally, this is just a teaser on some multi payload options that we have. This is probably next gen stuff, particularly given the rise of the site specific conjugation strategies that might have inherent limitations on the available number of conjugation sites.

The products on top are the building blocks, and you can put your own reactive groups on. The products in the middle are the ready to use cross linkers, and the product on the bottom is a potential custom cross linker for a dual MOA ADC. The chemist in the audience will recognize that there’s some orthogonal protection, deprotection steps. So it takes a minute or two longer, but this is certainly doable.

So to conclude, off target toxicity is one of the major obstacles to developing a clinically successful ADC. While it remains really difficult to predict the final safety profile and therapeutic window, there are a variety of preclinical assays that can be used to assess these potential downstream tox risks. And strategic linker payload design is demonstrated in here, has a track record for improving ADC performance in these assays. Many of the underlying mechanisms for tox are related to instability and nonspecific uptake, uptake, and they can be influenced by the physical chemical properties of the conjugate and of the linker payload.