Phenotypic plasticity, the ability of a living organism to respond to the environment, can lead to conclusions from experiments that are idiosyncratic to a particular environment. The level of environmental responsiveness can result in difficulties in reproducing studies from the same institute with the same standardised environment. Here we present a multi-batch approach to in-vivo studies to improve replicability of the results for a defined environment. These multi-batch experiments consist of small independent mini-experiments where the data are combined in an integrated data analysis to appropriately assess the treatment effect after accounting for the structure in the data. We demonstrate the method on two case studies with syngeneic tumour models which are challenging due to high variability both within and between studies. Through simulations and discussions, we explore several data analysis options and the optimum design that balances practical constraints of working with animals versus sensitivity and replicability. Through the increased confidence from the multi-batch design, we reduce the need to replicate the experiment, which can reduce the total number of animals used.The replicability crisis is a methodological crisis in science where results of many scientific studies are difficult to replicate 1 . This crisis is undermining preclinical research and raises ethical implications over the continued use of animals if the outcomes are questionable 2 . The reasons are multi-faceted, from incentive structure of science to methodological issues in the analysis and design. Research involving living organisms has an additional challenge to replicability in that living organisms are highly responsive to the environment with phenotypic changes with both long-and short-term durations. This ability, described as phenotypic plasticity, is an evolutionary adaptation to ensure optimal fit to the environment and an essential component for survival.The sensitivity of in-vivo experiments and concerns over the inability to replicate studies resulted in a call to increase standardisation and subsequent reporting 3-5 . It was felt that experiments with standardisation of procedures, environmental conditions, genetic background and tested where possible within one day would minimise sources of variation. Therefore, these were proposed as the gold standard approach as this approach would reduce variability, thereby enhance sensitivity and supports reproducibility when the environment is defined precisely. However, in a pivotal study, scientists went to great lengths to standardise the environment and protocols in a characterisation of multiple mice strains within three laboratories with several behavioural screens. Despite this extensive standardisation, they observed disparate results and proposed that interactions between the genotype and local environment led to idiosyncratic phenotypes 6 . The initial reaction was a call to increase the effort to standardise 7 .The level of responsiveness to the environment, had led ...

A proteolysis-targeting chimera (PROTAC) is a new technology that marks proteins for degradation in a highly specific manner. During screening, PROTAC compounds are tested in concentration–response (CR) assays to determine their potency, and parameters such as the half-maximal degradation concentration (DC50) are estimated from the fitted CR curves. These parameters are used to rank compounds, with lower DC50 values indicating greater potency. However, PROTAC data often exhibit biphasic and polyphasic relationships, making standard sigmoidal CR models inappropriate. A common solution includes manual omitting of points (the so-called masking step), allowing standard models to be used on the reduced data sets. Due to its manual and subjective nature, masking becomes a costly and nonreproducible procedure. We therefore used a Bayesian changepoint Gaussian processes model that can flexibly fit both nonsigmoidal and sigmoidal CR curves without user input. Parameters such as the DC50, maximum effect Dmax, and point of departure (PoD) are estimated from the fitted curves. We then rank compounds based on one or more parameters and propagate the parameter uncertainty into the rankings, enabling us to confidently state if one compound is better than another. Hence, we used a flexible and automated procedure for PROTAC screening experiments. By minimizing subjective decisions, our approach reduces time and cost and ensures reproducibility of the compound-ranking procedure. The code and data are provided on GitHub ( https://github.com/elizavetasemenova/gp_concentration_response ).

Anti-hormonal therapy has been the mainstay treatment for estrogen receptor (ER) +ve breast cancer for >40 years. The selective ER degrader (SERD) fulvestrant has demonstrated clinical benefit over aromatase inhibitors and the selective ER modulator tamoxifen. However, fulvestrant's efficacy may be limited by the intramuscular route of administration. To overcome this, AstraZeneca has developed orally bioavailable SERDs, the first generation compound AZD9496 and the next generation oral SERD AZD9833. Here we report the preclinical comparison of fulvestrant, AZD9496 and AZD9833. All three compounds induced equal ERα degradation in the ER+ breast cancer cell line MCF7. Importantly, analysis of other ER+ cell lines revealed AZD9833 phenocopied ERα degradation induced by fulvestrant, while AZD9496 only reduced ERα levels to 54% of that achieved by fulvestrant. Activity in the endometrial Ishikawa cell line demonstrated induction of progesterone receptor, an ER target gene, by AZD9496 but not AZD9833. RNAseq analysis of ER+ breast cancer cell lines treated with the three agents with and without estradiol did not reveal differences between the compounds; all estradiol-induced gene expression changes were completely reversed by the SERDs and no ER agonism was detected. By contrast, AZD9833 and fulvestrant produced an equivalent maximal anti-proliferative effect in MCF-7 and CAMA-1 cells, while AZD9496 induced a significantly inferior anti-proliferative effect. The pure anti-estrogen nature of AZD9833 and the partial agonist potential of AZD9496 was confirmed in vivo in a juvenile rat uterine assay. Like fulvestrant, AZD9833 caused a significant reduction in uterine weight and endometrial thickness, while AZD9496 produced a significant increase in both parameters. In MCF7 xenografts, both oral SERDs caused equivalent anti-tumour and pharmacodynamic effects to a supraclinical exposure of fulvestrant. However, in the patient derived xenograft CTC174 harbouring an ESR1 D538G mutation, AZD9833 caused equivalent anti-tumour and ERα degradation effects to supraclinical fulvestrant. By contrast, AZD9496 produced significantly poorer maximal anti-tumour and pharmacodynamic effects in this model compared to fulvestrant. These data demonstrate that AZD9833 phenocopied fulvestrant preclinically, as a SERD and pure anti-estrogen. It is, therefore, a suitable candidate to test the hypothesis that increasing the exposure of a SERD above what is achieved with fulvestrant can provide additional patient benefit. AZD9496, on the other hand, demonstrated inferior ERα degradation in several ER+ breast cancer cell lines and partial agonism. This may explain the result of the clinical window of opportunity study, comparing AZD9496 to fulvestrant, in which AZD9496 failed to demonstrate superior degradation of ERα, downregulation of PR expression and reduction in Ki67 positivity. Citation Format: Mandy Lawson, Natalie Cureton, Michelle DuPont, Oona Delpuech, Dawn Trueman, Pei Zhang, Azadeh Cheraghchi-Bashi-Astaneh, Sladjana Gagrica, Gareth Maglennon, Daniel Sutton, Bairu Zhang, Jonathan Cairns, Jamie Scott, Teresa Klinowska, Christopher J. Morrow. Not all selective estrogen receptor degraders are equal - Preclinical comparison of AZD9833, AZD9496 and fulvestrant [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 4379.

Anti-hormonal therapy has been the mainstay for the treatment of estrogen receptor (ER) positive breast cancer for over 40 years. The selective ER degrader (SERD) fulvestrant (Faslodex) has been shown to provide clinical benefit over aromatase inhibitors and selective ER modulators. However, delivery of its maximum possible efficacy may be limited by the intra muscular route of administration. To overcome this, AstraZeneca developed the orally bioavailable SERD AZD9496. Here we report further preclinical characterisation of AZD9496 with regards to its ability to induce ERα degradation, antagonise and agonise ER in vitro and the impact on tumour xenografts. AZD9496 caused equivalent ERα degradation to fulvestrant in the MCF7 ER+ cell line. Interestingly, in other ER+ cell lines (e.g. CAMA1 and T47D), the maximal level of ERα degradation induced by AZD9496 was 54% of that induced by fulvestrant. RNAseq analysis of cell lines treated with AZD9496 or fulvestrant in the presence of estradiol did not reveal any differences in ER antagonism. Furthermore, no ER agonism was detected in ER+ breast cancer cell lines treated with AZD9496 in the absence of estradiol. However, unlike fulvestrant, AZD9496 was able to induce PR protein expression in the Ishikawa ER+ endometrial cancer cell line, demonstrating a potential to agonise ER. Also, in proliferation assays, while there was little difference in MCF7 cells between the maximal antiproliferative effect of fulvestrant and AZD9496, in CAMA1 and T47D cells the maximal antiproliferative effect of AZD9496 was 75% and 82% respectively of that caused by fulvestrant. In vivo, AZD9496 increased the uterine weight and induced PR expression in the uterus of juvenile rats, consistent with AZD9496 being a partial ER agonist albeit less than tamoxifen. AZD9496, at a maximally efficacious dose, did not induce the same degree of ER degradation or anti-tumour effect as a supraclinical dose of fulvestrant in CTC174, a patient derived xenograft that harbours a D538G ESR1 mutation and grows in the absence of supplemental estradiol. Taken together, these data suggest that, while AZD9496 is equivalent to fulvestrant in terms of ERα degradation and anti-proliferative/anti-tumour effects in some preclinical models, this is not the case in all models. However, it is not clear which, if any, model accurately reflects the clinical setting and how AZD9496 would compare to fulvestrant at exposures achievable in the clinic. Thus, testing AZD9496 vs fulvestrant in ER+ breast cancer patients is crucial to determine potential clinical benefit. Citation Format: Mandy Lawson, Natalie Cureton, Oona Delpuech, Pei Zhang, Azadeh Cheraghchi-Bashi-Astaneh, Sladjana Gagrica, Dawn Trueman, Gareth Maglennon, Daniel Sutton, Bairu Zhang, Jonathan Cairns, Teresa Klinowska, Christopher J Morrow. Oral selective estrogen receptor degrader AZD9496 demonstrated preclinical model specific differences to fulvestrant in estrogen receptor degradation, agonism and anti-tumour effects [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P6-04-15.

Introduction: Microphysiological systems (MPS; organ-on-a-chip) aim to recapitulate the 3D organ microenvironment and improve clinical predictivity relative to previous approaches. Though MPS studies provide great promise to explore treatment options in a multifactorial manner, they are often very complex. It is therefore important to assess and manage technical confounding factors, to maximise power, efficiency and scalability.Methods: As an illustration of how MPS studies can benefit from a systematic evaluation of confounders, we developed an experimental design approach for a bone marrow (BM) MPS and tested it for a specified context of use, the assessment of lineage-specific toxicity.Results: We demonstrated the accuracy of our multicolour flow cytometry set-up to determine cell type and maturity, and the viability of a “repeated measures” design where we sample from chips repeatedly for increased scalability and robustness. Importantly, we demonstrated an optimal way to arrange technical confounders. Accounting for these confounders in a mixed-model analysis pipeline increased power, which meant that the expected lineage-specific toxicities following treatment with olaparib or carboplatin were detected earlier and at lower doses. Furthermore, we performed a sample size analysis to estimate the appropriate number of replicates required for different effect sizes. This experimental design-based approach will generalise to other MPS set-ups.Discussion: This design of experiments approach has established a groundwork for a reliable and reproducible in vitro analysis of BM toxicity in a MPS, and the lineage-specific toxicity data demonstrate the utility of this model for BM toxicity assessment. Toxicity data demonstrate the utility of this model for BM toxicity assessment.