AbstractDespite major advancements in the treatment of many cancers, pancreatic cancer has one of the poorest survival rates of all, with a ten-year survival of less than 1%. One experimental clinical approach to treat pancreatic cancer had involved targeting an apoptosis pathway, the TNF-related apoptosisinducing ligand (TRAIL) receptor 2/ death receptor-5 (DR5) pathway using the monoclonal antibody conatumumab (AMG 655). However, the results of these clinical trials proved disappointing with no survival advantage being observed. Subsequent literature has suggested that this was most likely due to inadequate receptor clustering and consequent lack of tumour cell apoptosis. In view of this, this thesis attempted to address the lack of efficacy with conatumumab through developing a nanoparticle-based approach armed with an entrapped chemotherapeutic and driving receptor clustering by decorating the nanoparticles with conatumumab.
The initial work in this thesis sought to test various nanoformulation approaches with the aim of characterising size and manufacturability, entrapment, ligand conjugation and in vitro efficacy. Initially, a formulation method was developed to generate nanoparticles of smaller size than traditionally formulated from the single emulsion method. A hybrid method combining salting out and single emulsion nanoparticle formulation techniques proved better than the single emulsion method in reducing particle diameter. However, with this method entrapment proved inadequate achieving only 1% efficiency. Entrapment of two established chemotherapy drugs was attempted including gemcitabine and camptothecin, and an experimental FLIP inhibitor 3642. Given its hydrophilic nature, a derivative of the double emulsion method was employed to entrap gemcitabine. But, while this showed in vitro efficacy, concerns regarding entrapment and ligand conjugation quantification precluded further development. A derivative of the single emulsion method was used to entrap 3642 and in vitro efficacy was achieved, this was not taken forward due to restrictions with freedom-to-operate and a lack of drug substance.
The topoisomerase 1 inhibitor camptothecin was chosen as the payload of choice in the nanoparticle candidate going forward. Camptothecin had a number of advantages 1) it showed favourable entrapment using the single emulsion method given its hydrophobic nature, 2) it is fluorescent, meaning entrapment could be quantified with ease, 3) the much less potent topoisomerase 1 inhibitor, irinotecan is currently used in pancreatic cancer treatment as part of the FOLFIRINOX regimen.
To improve antibody conjugation a novel N-hydroxysuccinamide endcapped polymer was employed to circumvent the need for in situ EDC functionalisation of carboxyl groups. This resulted in improved AMG 655 conjugation efficiency versus previously observed methods with fewer manufacturing steps required.
In chapter 4 the functional potency of the nanoparticle was investigated using a number of pancreatic tumour cell lines. Initially the level of DR5 receptor expression was determined both by flow cytometry and by Western blot. Variable levels of DR5 expression were observed ranging from 15% in PANC1 to 75% in AsPC-1 cell lines.
Conjugation of AMG 655 to the nanoparticle surface (in the absence of drug) rendered it capable of reducing cell viability in all 4 cell line models treated. Notably this reduction did not correlate with DR5 surface expression but was consistent with observations in the literature. The reduction of viability was promising and the addition camptothecin improved this potency. The combined effect of conjugated AMG 655 and entrapped camptothecin reduced cell viability in a synergistic manner.
In Chapter 4 the role of inhibitors of apoptosis FLICE-inhibitory protein (FLIP) and inhibitor of apoptosis protein (XIAP) were explored. It was clear that camptothecin downregulated FLIP and XIAP in a dose dependent manner. To gauge whether this downregulation was a potential explanation for the synergy observed or an off-target effect, siRNA knockdown of FLIP splice forms and XIAP was employed to simulate the downregulation observed from camptothecin. This revealed that FLIP downregulation contributed significantly by enhancing the potency of conjugated AMG 655. Furthermore, CRISPR (clustered regularly interspaced short palindromic repeats) targeting of FADD (Fas Associated Via Death Domain) and caspase 8 revealed that the synergy observed using this formulation was DR5 pathway-dependent.
Given the success of testing in vitro, in vivo efficacy of the formulation was next investigated. Initially, cell line xenografts of MIA PaCa-2 and PANC-1 were employed and the results revealed that, when injected intravenously, the formulation reached the tumour site in quantities sufficient to block tumour growth. Patient-derived xenograft (PDX) models were next explored as a means to more accurately simulate the pancreatic cancer environment as these models form stroma and have not been immortalised. Unfortunately, no efficacy was observed in either PDX model. This may be due to the stromal barrier, a lack of DR5 expression or slower PDX growth compared to the cell line xenograft - requiring an altered dosage regimen. Moreover, the lack of histology or access to tumour samples that could be analysed for drug infiltration or apoptotic markers prevented appropriate evaluation of these models.
In summary, a lead nanoparticle candidate was developed exhibiting good manufacturability and cytotoxicity. The potent cytotoxicity observed was synergistic, associated with FLIP downregulation and was DR5 pathwaydependent. This formulation exhibited promising in vivo efficacy in cell line xenografts. Despite the lack of efficacy observed in the PDX models CPT DR5 NP has demonstrated potent cell kill and in vivo efficacy and given the lack of available therapeutic options in the clinic this formulation could be optimised for further evaluation in both in vitro and in vivo models.
|Date of Award||Jul 2020|
|Sponsors||Medical Research Council & National Institutes of Health|
|Supervisor||Daniel Longley (Supervisor) & Christopher Scott (Supervisor)|
- Death Receptor 5