Abstract 3989: Patient derived cancer organoids as a model to predict clinical response and drug discovery in colorectal cancer

Abstract Background: Patient derived cancer organoids (PDCOs) have shown to preserve cell morphology and better recapitulate patients’ cancers. Due to these advantages, PDCOs could be used as a model to better predict how a patient will respond to drug treatments. STX-478, a PIK3CAH1047R mutant specific inhibitor, has shown promise to circumvent severe toxicities which could aid in the tolerability of novel combinations. Here, we examine the impact of PDCOs for predicting patient response to chemotherapy and in drug discovery. Methods: PDCO lines were acquired from surgery and biopsies following consent and approved Institutional Review Board protocols. Seven PDCOs were imaged (Day 0) and treated for 48 hours based on the patient’s clinical treatment. Media was subsequently replaced and PDCOs re-imaged 48 hours later (Day 4). Diameter analysis was performed using ImageJ and percent relative change (PRC) and Glass’s Delta (GD) was determined for all treatments. Based on previous data, a Glass’s Delta over 1.25 has been shown to correlate with partial response, and under 0.75 correlates with progressive disease. Clinical response was determined based on MRI scans and RECIST v1.1 guidelines. In the drug screen, a CRC PDCO line with a PIK3CAH1047R mutation, was treated with 115 different inhibitors in combination with STX-478 for 48 hours. All plates were imaged on Day 0 and Day 2; diameter analysis was performed, and GD was determined for all treatment groups. Results: Seven PDCO lines were treated with FOLFOX or FOLFIRI and a GD range of 0.01-1.61 with a median of 0.75 were observed. Based on previously described GD sensitivity thresholds, two PDCO lines were responsive, and five were nonresponsive to treatment. The five non-responsive PDCO lines were all collected from the residual tumor after treatment with FOLFOX. FOLFIRI treatment of PDCO line CRC1 caused a significant response compared to all treatment groups (PRC= 22.4%, GD= 1.26, control: 70.3%, 5FU: 33.5%, 0.97, SN38: 43.1%, 0.71). Clinically, the patient had a 42% reduction in cancer size per RECIST v1.1, which is considered a partial response. In the STX-478 drug screen, six different combinations resulted in an enhanced response compared to single agents, including enasartinib HCl (EHCL; STX-478 single agent PRC= 29.1%, EHCL= 38.6%, and combination= -2.06%, STX-478 vs combination GD: 1.57; EHCL vs combination: 2.07). Additionally, NSC23766, a Rac1 inhibitor, also demonstrated enhanced response in combination (STX-478 single agent PRC= 29.1%, NSC23766= 38.0%, and combination= 9.2%, STX-478 vs combination GD: 1.00; NSC23766 vs combination: 1.42). Conclusion: PDCOs show exciting potential as a predictive model for patient response and drug discovery. These findings need to be validated in further studies. Citation Format: Alexa E. Schmitz, Shirsa Udgata, Maggie O. Smith, Rian Engeldinger, Yousef Gadalla, Xingqi Shen, Addison T. Zick, Alec B. Cornelio, Cheri A. Pasch, Dustin A. Deming. Patient derived cancer organoids as a model to predict clinical response and drug discovery in colorectal cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3989.