Tumor Microenvironments (TMEs) hold vital information that has the potential to inform successful drug development and reduce development failure rates from current 97%[1] and deliver effective therapies that have far higher response rates compared with current standard of care. Biosignature guided patient selection holds the key in improving success of clinical trials and reduce the estimated wastage of $ 60b[2] annually in failed clinical trials.

TME data also holds the potential for delivering precise functional information to predict clinical response by evaluating the impact of perturbation of tumor samples with existing therapies or their combinations and ascertain the likelihood of response, prior to subjecting patients to treatment.

 While the premise of TME research is exciting, there are several challenges in meaningfully utilizing human tumor microenvironments at scale, starting with accessing live human tumors and preserving them in a manner that they represent the original cancer in its entirety retaining their immune and stromal components intact. The value of studying dynamic signaling between all compartments of living tumors is immense, and labs across the world are engaged in research to decipher the intricacies of tumor micro-environments.

 With the platform’s combined experience of processing over 24,000 tumor samples over the course of a decade, Farcast has identified and overcome some of the challenges associated with processing live human tumors. Our optimized protocol ensures that tumors are harvested, collected and transported in a manner that ensures good quality starting material. Our fragmentation and distribution techniques involve delicate handling and minimal perturbation of tumors preserving the original architecture. By maintaining the tumor contexture, the platform has demonstrated functional fidelity of different sub-types of immune cell populations over a period of 72 hours. Our plating and culturing techniques recreate the heterogeneity of original tumors leading to multiple test arms for every tumor. The platform is agnostic to the nature of therapy and interrogates each tumor with multiple regimens simultaneously and offers readouts that compare the response across diverse treatment options. The application of this platform has helped several biotechs to elucidate the mechanism of action of their therapies during the translation phase and make informed development decisions with precision biosignatures to guide the development of their therapies. Biosignatures are particularly helpful in selection of patients who would best respond. Besides it opens up personalized treatment options with customized combinations rather than monotherapy, enabling a superior probability of success.

 A formidable patient sample size (n=98) was employed in the standardization and validation of the Head and Neck platform.  Multiple assay readouts from 70 patient samples were correlated to establish the activity of Nivolumab in live tumor fragments. 

The platform unequivocally confirmed via orthogonal data, the activity of Nivolumab as an effective check point inhibitor that led to statistically significant level of re-invigoration and activation of Cytotoxic T cells (CTLs) in the cohort. Given that clinical response rates are reported at a much lower level in HNSCC patients, our quest was to differentiate between “activity” and “efficacy” at a sample level. We used t-SNE based approach unbiased clustering based on 10 response parameters to clearly stratify responders and non-responders. While good responders (16%) exhibited both T-cell activity and enhanced tumor cytotoxicity, non-responders exhibited neither. Majority of the patient samples (73%) constituted the moderate responders.  We demonstrated significant T-cell infiltration into the tumor nest in good responders. A sub-set of moderate responders benefitted with the combination of Ipilimumab (anti-CTLA4) driven by distancing of inhibitory T-regs from the CTLs . The supportive role of NK cells and the negative role of macrophages on the efficacy of Nivolumab was also demonstrated in this study.

The study developed baseline and on-treatment gene expression signatures that could effectively stratify responders and non-responders in published clinical data sets. and confirmed to have predictive power. This deduction increases the confidence in the platform’s ability to develop biosignatures in future that will be useful in clinic as well as development of effective therapies.

Backstory:

 The idea of this publication was conceived during the lockdown period of COVID-19 and the team at Farcast persevered through 3 waves of the pandemic and stuck to its resolve of solving the problem of creating a novel biosignature that will help reduce the mortality and morbidity burden for HNSCC patients. The success of our study came by creating a cross-functional team of surgeons, medical oncologists and pathologists working in tandem to create a powerhouse of expertise in tumor biology and clinical acumen. The engagement of cancer centers ensured access to high quality tumors, and we are grateful for the patients who consented to donate their tumor samples for making this research possible.

 We remain committed to refining our signatures and have progressed to finding correlation between our prediction of response and resistance with clinical outcomes. We hope to publish a sequel to this publication after establishing concordance of our prediction.

 We hope that this paper will encourage other researchers to build on our findings and look forward to continuous engagement with the community of tumor biologists and Oncologists in the quest for furthering the field of precision functional Oncology.

[1] Biostatistics 2019, Wong et al

[2] BMJ 2020, Moore et al