Pathway-selective strategies reveal specific PI3K–mTOR inhibitors as key partners in KRAS G12D-targeted therapy

Pathway-selective strategies reveal specific PI3K–mTOR inhibitors as key partners in KRAS G12D-targeted therapy | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Pathway-selective strategies reveal specific PI3K–mTOR inhibitors as key partners in KRAS G12D-targeted therapy Sofie Seghers, Maxim Le Compte, Felicia Rodrigues Fortes, Geert Roeyen, and 10 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7614104/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract KRAS G12D is the most common oncogenic mutation in pancreatic ductal adenocarcinoma (PDAC), driving resistance and heterogeneity. Using robotic high-throughput drug screening with live-cell imaging, we evaluated normalized organoid growth rate (NOGR) in KRAS G12D-mutant PDAC organoids treated with the inhibitor MRTX1133. Monotherapy induced predominantly cytostatic, dose-dependent effects, reflecting heterogeneity observed in single-cell transcriptomics, where compensatory MAPK/ERK and PI3K–mTOR activation emerged. Drug response profiles were variable, and synergy screening revealed patient- and compound-specific interactions. The most consistent cytotoxic synergy was achieved with PI3K–mTOR inhibitors. Isoform-specific PI3Kα inhibitors (inavolisib, alpelisib) demonstrated robust synergy with MRTX1133 and favorable tumor selectivity, whereas dual PI3K/mTOR inhibitors (e.g., VS-5584) were more cytotoxic but lacked specificity. These results indicate that KRAS G12D inhibition alone is insufficient due to underlying transcriptional diversity, and highlight isoform-specific PI3Kα inhibitors as promising partners for combination therapy in PDAC. Biological sciences/Cancer Biological sciences/Drug discovery Health sciences/Oncology Pancreatic ductal adenocarcinoma KRAS G12D MRTX1133 patient-derived organoids single-cell RNA sequencing combination therapies drug synergy PI3K–mTOR inhibition targeted therapy resistance functional precision oncology Background Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive and treatment-resistant solid tumors, with a five-year survival rate around 13% and limited improvement in outcomes over recent decades ( 1 ). The disease is typically diagnosed at an advanced stage and remains refractory to most systemic therapies ( 2 ). Even among the minority of patients eligible for surgical resection, recurrence is frequent and prognosis remains poor ( 3 ). Thus, there is an urgent need for more effective, personalized therapeutic strategies. Mutations in the KRAS oncogene occur in over 90% of PDAC cases ( 4 ), making it the most ubiquitous and well-established driver of pancreatic tumorigenesis. Historically considered “undruggable,” recent advances have led to the development of small molecules that directly target specific KRAS mutants. Notably, inhibitors against KRAS G12C, such as sotorasib and adagrasib, have demonstrated clinical efficacy in non-small cell lung cancer (NSCLC) and other malignancies ( 5 , 6 ). However, these inhibitors are not broadly applicable to PDAC, where KRAS G12D is the most common mutation, present in approximately 40% of patients ( 4 ). Emerging compounds such as MRTX1133, which specifically target KRAS G12D, have shown promise in preclinical studies ( 7 ). Yet, KRAS inhibition in PDAC often leads to disease progression and resistance in most patients ( 8 ). This resistance underscores a fundamental challenge in targeting KRAS: despite its central role in tumor initiation, PDAC cells frequently develop alternative or compensatory signaling dependencies that undermine the durability of single-agent therapies ( 9 ). One contributing factor to therapeutic resistance is the high degree of inter- and intra-tumoral heterogeneity observed in PDAC ( 10 ). To address this complexity, there is growing interest in functional precision medicine approaches that integrate real-time drug testing with molecular profiling to identify patient-specific vulnerabilities. Patient-derived organoids (PDOs) have emerged as powerful ex vivo models that retain the histological and genetic features of the original tumor while enabling high-throughput drug screening ( 11 ). Combined with advances in single-cell RNA sequencing (scRNA-seq) and pathway activity analysis, PDO platforms offer a unique opportunity to capture both phenotypic and molecular heterogeneity and guide the rational design of combination therapies ( 12 ). In this study, we used a multimodal strategy to dissect the functional and molecular heterogeneity of KRAS-mutant PDAC and identify effective combination strategies to enhance the limited efficacy of KRAS G12D inhibition. We first evaluated the therapeutic response of patient-derived organoids to MRTX1133 and uncovered variable, predominantly cytostatic effects. We then explored a publicly available scRNA seq database of PDAC patient samples, to explore underlying signaling diversity and identify actionable pathway dependencies at single cell level. Finally, we designed and screened a panel of drug combinations, based on the pathway analysis from scRNAseq data, revealing patient-specific and compound-specific synergies. Results KRAS G12D inhibition shows limited cytotoxic efficacy in ex vivo models To evaluate the therapeutic potential of KRAS G12D inhibition in pancreatic cancer, we treated four patient-derived organoid models harboring KRAS G12D (n=3) or G12V (n=1) mutations with increasing concentrations of the KRAS G12D-specific inhibitor MRTX1133. We selected 120h as the optimal timepoint for quantifying the growth rate–based drug response metric normalized organoid growth rate (NOGR), as the observed variability in organoid growth among vehicle controls justifies the use of growth rate–corrected metrics over traditional relative viability measures (Figure 1A). The NOGR metric provides an accurate assessment of drug response by distinguishing between cytostatic effects (NOGR between 1 and 0) and cytotoxic effects (NOGR between 0 and -1). Even at higher doses (up to 3 mM), G12D-mutant organoids showed only partial growth inhibition without significant loss of viability, indicated by a positive NOGR value (Figure 1B). This cytostatic effect was also visually confirmed, as shown in Figure 1C. PDAC002 (KRAS G12V) remained largely unaffected, confirming inhibitor specificity. However, G12D organoids began to recover and resume proliferation over time, indicating only a transient growth-inhibitory effect, as exemplified by PDAC044 (Figure 1D). Live-cell imaging at 3 μM MRTX1133 further supported this observation, revealing a cytostatic rather than cytotoxic response and subsequent growth recovery in PDAC044 organoids (Figure 1E). Collectively, these findings indicate that KRAS G12D inhibition can transiently suppress tumor growth but has limited therapeutic efficacy as monotherapy, aligning with an increasing number of drug combination clinical trials in KRAS mutant patients (NCT07020221, NCT0694976, NCT05382559,…). Rational combination therapy development through single-cell RNA-seq and pathway profiling Given the predominantly cytostatic response to KRAS G12D inhibition observed in patient-derived organoids (Figure 1), we sought to uncover co-targetable vulnerabilities that might convert this effect into a cytotoxic response. Specifically, we aimed to address two key questions: is KRAS pathway activity heterogeneously distributed across tumor cells within PDAC tumors and can single-cell transcriptomic data be used to rationally identify novel combination therapy strategies tailored to the diverse signaling dependencies present in these tumors. To this end, we explored a publicaly available scRNA-seq dataset of PDAC tumors using Scanpy (13, 14). After normalization and quality control, we annotated epithelial tumor cells based on canonical epithelial markers, including EPCAM and KRT19. UMAP visualization of the integrated datasets, each representing an individual patient (Figure 2A, left), revealed significant inter- and intra-patient heterogeneity, with epithelial cells scattered across multiple distinct clusters (Figure 2A, right), suggesting functional and phenotypic diversity within and between tumors. Differential expression analysis revealed cluster-specific marker genes (Figure 2B), underscoring transcriptional divergence among tumor cells, supporting the hypothesis that distinct subpopulations may underlie patient-specific co-targetable vulnerabilities. To further evaluate the functional heterogeneity, we calculated global pathway activity scores for each patient using gene sets from Hallmark, Reactome, and WikiPathways (Figure 2C). This analysis showed clear differences in signaling pathway activation between patients, including variability in cell cycle control, stress response, and growth factor signaling. Next, we examined KRAS-specific signaling activity at single cell level using AUCell. As shown in Figure 2D, HALLMARK_KRAS_SIGNALING_UP scores varied widely across tumor clusters, confirming that KRAS-driven transcriptional programs are not uniformly active, even within tumors harboring KRAS mutations. In contrast, HALLMARK_KRAS_SIGNALING_DN scores were consistently low and uniformly distributed across all clusters, suggesting a broadly repressed transcriptional program that is typically downregulated when KRAS signaling is active. This pattern implies that while KRAS-driven oncogenic programs may differ in amplitude between subpopulations, a baseline level of KRAS activity is maintained across the tumor. Such a profile may reflect a minimal threshold of KRAS dependency common to all malignant epithelial cells, with superimposed heterogeneity arising from additional pathway activations. Similarly, transcriptional activity of tyrosine kinase pathways (REACTOME_SIGNALING_BY_RECEPTOR_TYROSINE_KINASES) also showed a rather ‘patchy’ distribution, indicating that kinase signaling dependencies differ at the single-cell level. To capture these functional differences in a more integrated manner, we performed PCA on global pathway scores and identified five distinct tumor subtypes (Figure 2E). Clustering of these subtypes revealed that each was characterized by unique pathway activation profiles based on their top 7 activated pathways (Figure 2F), including differential enrichment of MYC targets, E2F targets, EGFR, FGFR, and MAPK/ERK signaling, all pathways with known roles in PDAC progression and resistance. Based on these findings, we rationally designed a targeted drug panel (Figure 2G) that reflects the dominant signaling axes active within the different tumor subpopulations. This strategy offers a path toward more effective and personalized combination therapies by directly addressing the observed functional heterogeneity. In summary, our single-cell transcriptomic analysis lays the groundwork for rationally designed combination treatments with KRAS inhibitors, tailored to the specific signaling dependencies of PDAC subtypes. Kinase inhibitor screening confirms functional heterogeneity predicted by single-cell transcriptomics Based on the signaling pathway heterogeneity identified in our single-cell transcriptomic analysis (Figure 2D), we next examined whether this patient-specific heterogeneity was also captured in our organoid panel and whether it could be linked to differences in drug response. Geneset signature maps were generated using UMAP clustering of genes by covariance, with coloring based on relative (average) log-expression in the phenotype group (Figure S1, supplementary table 1). This projection highlights geneset co-expression patterns and underscores the marked heterogeneity in pathway activity across organoid lines. Figure 3A provides an overview of selected genesets relevant to the drug panel shown in Figure 2G, revealing enhanced enrichment of these pathways in PDAC060 compared to the other organoid lines. Notably, a similar sample-level clustering pattern was observed in the drug response data, where PDAC060 exhibited marked hypersensitivity to the majority of tested compounds, as reflected by the Area Over the Curve (AOC) of the NOGR dose-response curves (Figure 3B, Figure S2). The MEK inhibitors trametinib (selective) and ASTX029 (dual MEK/ERKi) showed the strongest anticancer activity across all organoid lines, with PDAC060 being the most sensitive line (Figure 3B). A detailed compound list is provided in Supplementary Table 2. To quantify the extent of response variability, we ranked all tested compounds by the standard deviation of their AOC values across organoid lines (Figure 3C). Compounds such as sitravatinib, and trametinib showed a great degree of variability, suggesting they may be particularly influenced by tumor-specific signaling contexts. Finally, a Pearsons correlation analysis comparing drug responses to MRTX1133 revealed both overlapping and distinct response patterns (Figure 3D), providing insight into potential combinatorial strategies or resistance mechanisms. Together, these data confirm that the signaling heterogeneity observed at the transcriptomic level is also observed in the differential drug sensitivities. This reinforces the rationale for personalized combination therapies targeting dominant pathways within each tumor subtype. Synergy screening identifies MAPK–ERK and PI3K–mTOR axis combinations as promising therapeutic strategies in PDAC To build on the observed pathway heterogeneity, we screened rational drug combinations in four KRAS-mutant PDAC organoid models, focusing on inhibitors of MAPK, PI3K–AKT–mTOR, and related kinase pathways. This to evaluate whether dual targeting could overcome the cytostatic limitations observed with KRAS G12D monotherapy. To assess whether observed synergy was specifically dependent on KRAS G12D inhibition or could also arise through parallel pathway interactions, we included a subset of combinations using paxalisib (a dual PI3K–mTOR inhibitor) as the backbone compound. This allowed us to decouple synergy from direct G12D targeting and evaluate whether convergent inhibition of survival signaling (e.g., PI3K–mTOR + MAPK) could produce similar or stronger effects in models with alternative KRAS mutations, such as PDAC002 (KRAS G12V). Synergy between drug pairs was evaluated using a 3x3 synergy matrix and quantified using four established reference models: Highest Single Agent (HSA), Loewe, Bliss, and ZIP. As shown in Figure 4A, we observed a high degree of correlation between the HSA and Loewe models (Pearson r > 0.95), suggesting strong agreement in synergy classification across these two metrics. Similarly, the ZIP and BLISS models correlated most with each other. Violin plots of synergy scores across all combinations (Figure 4B) revealed that higher synergy scores were obtained with HSA and Loewe, consistent with previous reports that ZIP and BLISS may be more conservative in detecting synergy. We selected HSA for further downstream analysis, focusing on combinations that outperform the most effective single agent, as it provides a more interpretable and less conservative estimate of synergy compared to ZIP and BLISS models. All individual synergy matrices for each unique drug-organoid combination are provided in Supplementary File 1. Figure 4C summarizes the mean HSA synergy scores for drug combinations with 3 μM MRTX1133. The bubble plot visualizes both the level of synergy (bubble size and category) and the corresponding NOGR values (color gradient), allowing distinction between cytostatic (blue, large bubble) and cytotoxic (red, large bubble) synergistic interactions. Consistent with the monotherapy responses, we observed organoid-specific synergistic drug combinations. As expected, no synergistic interactions with MRTX1133 were detected in PDAC002, which lacks the KRAS G12D mutation. In contrast, PDAC060, characterized by the strongest pathway activation among all organoid lines, exhibited the highest number of synergistic interactions across nearly all drug classes. Synergy in PDAC044 was particularly enriched for PI3K pathway-targeting drugs. PDAC087, on the other hand, was the most resistant, showing only limited cytotoxic synergy. For drug combinations involving the dual PI3K–mTOR inhibitor paxalisib, synergy was primarily observed with MEK inhibitors such as trametinib and ASTX029. Notably, this included organoid lines like PDAC087, for which few synergistic combinations with MRTX1133 were identified, highlighting a potential alternative strategy. Overall, drug combinations targeting the PI3K/mTOR axis had the broadest activity in both PDAC044 and PDAC060. Therefore we selected this therapeutic class to expand the targeted drug panel. Synergy with MRTX1133 is driven by chemical structure differences and not solely by target class Building on earlier observations that the PI3K/mTOR pathway represents a promising axis for combination therapy in KRAS G12D-mutant PDAC, we next aimed to dissect the compound-specific and contextual factors that shape synergy with MRTX1133. To explore this, we assembled a chemically diverse panel of 34 inhibitors targeting the PI3K/mTOR axis and tested these in a 4x1 synergy matrix (1 mM MRTX1133). This included isoform-specific PI3K inhibitors (e.g. taselisib, inavolisib), dual PI3K/mTOR inhibitors (e.g. omipalisib, gedatolisib) and pan-PI3K inhibitors (e.g. Pictilisib, AMG511). A detailed compound list is provided in Supplementary Table 2. All compounds were tested in combination with MRTX1133 across six cell models, including KRAS G12D- and G12V-mutant PDAC organoids, a KRAS G13D cholangiocarcinoma model (CCA012), and non-malignant lung organoids (LU046). This experimental setup enabled the prioritization of drug combinations that selectively target KRAS G12D mutant cancer cells while sparing healthy epithelial cells. Figure 5A shows a clustered heatmap of monotherapy drug responses across organoid lines, based on the AOC of NOGR dose–response curves (Supplemental Figure 3). Hierarchical clustering revealed distinct response patterns, with the most potent compounds clustering at the top, predominantly consisting of dual PI3K/mTOR inhibitors. Notably, healthy epithelial lung organoids (LU046) clustered closely with the most sensitive cancer organoid lines PDAC060 and CCA012, suggesting that these therapeutic classes lack selectivity for cancer cells in this conventional organoid model system. Figure 5B summarizes the mean HSA synergy scores for drug combinations with 1 μM MRTX1133 and the corresponding NOGR values. Detailed synergy matrices are provided in Supplementary File 2. Synergistic effects were specifically observed in KRAS G12D mutant organoid lines, but were absent in KRAS G13D (CCA012), KRAS G12V (PDAC002), and healthy lung organoids (LU046). This selective synergy was evident with both PI3K inhibitors and dual PI3K/mTOR inhibitors, supporting the notion that effective combinations with MRTX1133 can be tumor-selective, and that synergy is not the result of additive pathway inhibition, but a result of tumor-specific dependencies and drug-structure interactions. To identify tumor selective drug combinations, we filtered for inhibitors with an NOGRmax > 0.5 (effect a 3 mM) in LU046, indicating a marginal growth inhibitory effect in healthy epithelial cells. Of the 10 inhibitors that met these criteria (Supplemental Figure 4A), only AZD8186 showed selective synergy: cytostatic in PDAC044, cytotoxic in PDAC060, and additive cytostatic in PDAC087 (Figure 5B (green), Figure 5C-D). As monotherapy, all organoid lines exhibited a predominantly cytostatic response profile (Figure 5E). Notably, only apitolisib, pictilisib, and VS-5584 induced substantial synergistic effects across all three KRAS G12D-mutant organoid lines (Figure 5B (red), Figure S4B–D). Among these, VS-5584 was the most potent, inducing a strong synergistic response at 300 nM in combination with 1 μM MRTX1133, as illustrated by the live-cell images (Figure S4E). However, VS-5584 also showed high monotherapy potency across all organoid lines, including the healthy epithelial lung organoids LU046 (Figure S4F), raising concerns about selectivity. Of the four FDA-approved PI3K inhibitors tested (alpelisib, idelalisib, duvalisib and inavolisib), only alpelisib and inavolisib demonstrated synergistic interactions with MRTX1133 (Figure 5B, blue). Inavolisib was the most potent, showing consistent synergy across all tested concentrations in PDAC044 and PDAC060 (Figure S4G-H). Notably, inavolisib induced only a cytostatic effect as monotherapy in all organoid lines, including the healthy epithelial lung organoids LU046N. Figure 5B highlights the compound- and model-specific nature of these responses. While omipalisib, voxtalisib, and PQR530 are all dual PI3K–mTOR inhibitors, only omipalisib and PQR530 elicited synergy in specific models, whereas voxtalisib showed negligible activity, reinforcing that nominal target overlap does not predict functional synergistic outcome. In summary, these data reveal that monotherapy efficacy and synergy with KRAS G12D inhibition is not solely determined by PI3K–mTOR target class, but rather by a combination of tumor-intrinsic signaling dependencies and compound-specific properties. Even among agents with similar reported targets, response profiles diverge significantly between models, reinforcing the need for chemotype-informed combination strategies that account for intertumoral heterogeneity and potential off-target effects of TKIs. Discussion Recent advances in the development of KRAS G12D inhibitors have re-energized efforts to directly target mutant KRAS in PDAC. MRTX1133 represented the first clinical inhibitor of KRAS G12D, with preclinical studies demonstrating significant tumor regression in KRAS G12D–mutant models ( 15 ). However, early clinical evaluation revealed modest and transient responses in patients, compounded by suboptimal pharmacokinetic properties, ultimately halting further development. These limitations reflect not only the pharmacological challenges inherent to targeting KRAS G12D but also the plasticity and heterogeneity of tumor signaling networks. KRAS dependency in PDAC is neither binary nor uniform. Even within KRAS-mutant tumors, we observe substantial heterogeneity in reliance on canonical KRAS-driven pathways ( 16 ). Single-cell RNA sequencing analysis of KRAS G12D–mutant epithelial cells revealed uneven KRAS signaling activity, with distinct subpopulations exhibiting activation of alternative programs, including receptor tyrosine kinase signaling, MYC-driven transcriptional targets, and cellular stress response pathways. This intratumoral heterogeneity enables dynamic rewiring of signaling dependencies, facilitating adaptive resistance to KRAS G12D inhibition. Consistent with this, MRTX1133 monotherapy elicited predominantly cytostatic responses in vivo, likely reflecting the engagement of compensatory survival pathways upon KRAS inhibition. These findings underscore a fundamental limitation of single-agent targeted therapies in PDAC: inhibition of a single oncogenic node within a highly adaptive signaling network is insufficient to achieve durable tumor control. KRAS activating mutations drive strong dependency on the PI3K/AKT pathway ( 17 ). Ihle et al. showed that KRAS G12D mutateted NSCLC models displayed elevated PI3K and MEK/ERK signaling ( 18 ), and pharmacological and CRISPR-based screens identified PI3Kα as a critical bypass pathway whose co-inhibition with KRAS G12D enhances anti-tumor activity ( 19 ). Consistently, work with the inRas37 antibody combined with the PI3K/mTOR inhibitor BEZ-235 in pancreatic cancer showed that dual inhibition of RAS/MAPK and PI3K pathways effectively blocked MAPK reactivation, significantly suppressing tumor growth in vitro and in vivo ( 20 ). To address this, we implemented rational combination strategies guided by functional drug screening and pathway activity inferred from single-cell transcriptomic data. Although patient-specific synergistic combinations were observed, such as MRTX1133 and ASTX029 in PDAC060, the most consistent synergy across patients was seen when MRTX1133 was combined with PI3K–mTOR pathway inhibition. These findings are consistent with a recent study by Misale and colleagues, which demonstrated that among all targets screened, KRAS (G12C) blockade in combination with PI3K pathway inhibition in NSCLC produced the most robust therapeutic effect ( 21 ). Among the PI3k-mTOR combination regimen tested, AZD8186 (a PI3Kb/d inhibitor) showed the most tumor-selective synergy, but remained largely cytostatic, inducing growth arrest rather than cell death. This highlights an important consideration in combination therapy design: tumor-selective combinations can mitigate off-target toxicities and improve tolerability, but their clinical benefit may be limited if the synergy does not translate into meaningful cytotoxicity. This challenge is illustred by the halted phase Ib/II trial of AZD8186 in combination in advanced gastric cancer due to futility ( 22 ). Nevertheless, a phase I trial combining AZD8186 with abiraterone or the mTOR inhibitor vistusertib reported preliminary anti-tumor activity ( 23 ). Preclinical studies in Pten;Trp53-null mice further demonstrated that AZD8186 combined with selumetinib, a MEK inhibitor, improved survival and reduced mesothelioma proliferation ( 24 ), supporting vertical co-targeting of the KRAS–MEK axis and PI3K signaling to overcome resistance. In contrast to the tumor selective AZD8186, apitolisib (PI3K/mTORi), pictilisib (a pan-PI3Ki) and VS5584 (PI3K/mTOR), emerged as a highly synergistic partners that, despite lower tumor selectivity, induced more pronounced tumor cell death in combination. These findings emphasize that synergy alone does not guarantee clinical success. Even statistically significant interactions may fail to achieve the absolute level of tumor control required for durable responses if neither agent demonstrates sufficient monotherapy efficacy ( 25 ). Of these three compounds, VS5584 was the most potent in KRAS G12D-mutant cells. While two clinical trials (NCT01991938, NCT02372227) were terminated early due to limited recruitment, preclinical data support its rational use. Notably, Ning et al. showed superior tumor control with VS-5584 plus an ERK inhibitor, underscoring the value of dual PI3K/mTOR and MAPK/ERK pathway blockade ( 26 ). These findings underscore that synergy alone is insufficient unless paired with meaningful cytotoxic efficacy and mechanistic plausibility, especially in settings where vertical pathway inhibition may be needed to overcome adaptive resistance. Apitolisib has completed 12 clinical phase I/II trials, showing modest but durable activity in solid tumors ( 27 ), though a trial in RCC found it less effective than everolimus ( 28 ). Across studies, severe on-target toxicities (mainly hyperglycemia) have limited its clinical potential ( 27 – 29 ). Pictilisib also shows promise: preclinical studies demonstrated synergy with siKRAS in KRAS G12D/PTEN-deficient models ( 30 ), but clinical trials, both as monotherapy ( 31 , 32 ) and in combination ( 33 – 35 ), have reported mixed results ( 31 – 37 ). While some trials reported acceptable tolerability with encouraging anti-tumor activity, others highlighted limited efficacy, often due to dose-limiting toxicities. Unlike the investigational PI3K inhibitors described above, alpelisib and inavolisib are clinically validated PI3Kα-selective inhibitors, offering a clear translational advantage for combination with KRAS G12D inhibitors. Both target PI3Kα-driven signaling, a key bypass mechanism upon KRAS inhibition ( 38 ). Alpelisib, FDA-approved for HR+/HER2– breast cancer with PIK3CA mutations (SOLAR-1 trial) ( 39 ), effectively suppresses PI3K-mediated feedback activation of RTK/MAPK pathways, supporting its rationale in dual KRAS–PI3K targeting. Inavolisib, a next-generation PI3Kα inhibitor with mutant PIK3CA degradation activity, has shown superior and durable pathway suppression, with recent INAVO120 data confirming clinical benefit in HR+/PIK3CA-mutant breast cancer ( 40 ). Both agents showed strong cytotoxicity in 2 of 3 KRAS G12D-mutant PDAC lines when combined with MRTX1133, contrasting with the cytostatic effects of AZD8186. Their FDA approval and advanced clinical data lower translational barriers for combination trials, though toxicity management remains critical. When comparing these classes of inhibitors, a key observation is the differential toxicity and activity of PI3K-targeting agents. Dual PI3K/mTOR inhibitors (e.g., gedatolisib, apitolisib, VS-5584) show the broadest activity, targeting multiple pathway nodes. However, this broad activity comes at the cost of greater toxicity, as also seen in our lung organoid models, making them less suitable for combination strategies and better suited for select single-agent use ( 41 ). Pan-PI3K inhibitors have a narrower activity range but still present systemic toxicities ( 41 ), whereas isoform-specific inhibitors (e.g., alpelisib, inavolisib) offer a more favorable safety profile and could be suited for combination therapy. Future work should focus on preclinical validation of these combinations in vivo, exploration of predictive biomarkers, and ultimately, clinical testing to determine whether KRAS–PI3K co-targeting can achieve durable responses in PDAC. Conclusion In summary, our study highlights the potential of combining KRAS G12D inhibition with PI3K pathway blockade to overcome the cytostatic limitations of MRTX1133 monotherapy. While dual PI3K/mTOR inhibitors show broad activity, their toxicity profile limits clinical translation, particularly in combination regimens. Conversely, isoform-specific inhibitors such as alpelisib and inavolisib not only demonstrated robust synergy with MRTX1133 in KRAS G12D-mutant PDAC models but also benefit from established clinical use and manageable safety profiles. These findings underscore the importance of rational inhibitor selection based on both tumor biology and tolerability. Methods Patient Material The patient-derived organoids used in this study are registered in the Biobank@UZA (Antwerp, Belgium; ID: BE71030031000); Belgian Virtual Tumorbank funded by the National Cancer Plan. Organoids were derived from resection fragments or biopsies obtained from cancer patients treated at the Antwerp University Hospital and a written informed consent was obtained from all patients. The study was approved by the UZA Ethical Committee (ref. 17/30/339 and 14/47/480). Organoid cultures Basic medium consisted of Ad-DF + ++ (Advanced DMEM/F12 (GIBCO, #12634028), with 1% GlutaMAX (GIBCO, 35050-061), 1% HEPES (GIBCO, #15630080), 1% penicillin/streptomycin (GIBCO, #15140122) supplemented with 2% Primocin (Invivogen, #ant-pm-05). For PDAC and CCA organoids, Ad-DF + ++ was supplemented with 0.5 nM WNT Surrogate-Fc-Fusion protein (ImmunoPrecise, #N001), 4% Noggin-Fc Fusion Protein conditioned medium (ImmunoPrecise, #N002), 4% Rspo3-Fc Fusion Protein conditioned medium (ImmunoPrecise, #R001), 1x B27 (Gibco, #17504044), 10 mM nicotinamide (Sigma-Aldrich, #N0636), 1.25 mM N-acetylcysteine (Sigma-Aldrich, #A9165), 100 ng/ml FGF-10 (Peprotech, #100 − 26), 500 nM A83-01 (Tocris, #2939), 10 nM gastrin (R&D Systems, #G9145) and 10 µM Y-27632 after passaging (Selleck Chemicals, #S1049). Normal pulmonary organoids were cultured in Ad-DF+++ supplemented with 4% Noggin-Fc Fusion Protein conditioned medium (ImmunoPrecise), 4% Rspo3-Fc Fusion Protein conditioned medium (ImmunoPrecise), 1x B27 (Gibco), 10 mM nicotinamide (Sigma-Aldrich), 1.25 mM N-acetylcysteine (Sigma-Aldrich), 100 ng/ml FGF-10 (Peprotech), 25 ng/ml FGF-7 (Peprotech), 500 nM A83-01 (Tocris) and 1 µM SB202190 (Sanbio, Cayman Chemical). For passaging, the organoids were digested to single cells with TrypLE Express (GIBCO, #12604021) and resuspended in > 80% ice cold Cultrex growth factor reduced BME type 2 (R&D Systems, #3533-010-02) in full organoid medium. Small droplets of 20 µL were plated and incubated inverted for 30 min at 37°C to allow them to solidify after which the drops were covered with full organoid medium. In-depth characterization of PDAC organoids used in this study is described in Le Compte et al., 2023 ( 42 ) and the normal pulmonary organoid LU046 in Deben et al., 2023 ( 43 ). Drug screening Drug screening on 3D organoids was performed at the DrugVision.AI automated screening facility of the University of Antwerp, Belgium, using a prevalidated drug screening pipeline for which a detailed protocol is available in the Journal of Visualized Experiments ( 44 ). Briefly, established organoid lines were expanded in ECM domes (Cultrex type 2, R&D Systems, #3533-010-02). Next, 3-day-old organoids were harvested from ECM drops using the Cultrex Organoid Harvesting Solution (R&D Systems, #3700-100-01), collected in a 15 mL tube coated with 0.1% BSA/PBS, washed, and resuspended in medium. Next, the number of organoids was quantified by adding 5 µL of the organoid solution to 45 µL of medium in a 384-microplate well. A whole-well brightfield image was captured using the Tecan Spark Cyto and the number of organoids was counted label-free using Orbits ® (Orbits Oncology). Next, the organoid solution was diluted in full medium supplemented with 4% Cultrex at a concentration of 4000 organoids / mL. Next, 50 µL (200 organoids) of this solution was dispensed into each well of a 384-well ultra-low attachment microplate (Corning, #4588) using the OT-2 pipetting robot (Opentrons) in a cooled environment. Thereafter, the plate was centrifuged (100 rcf, 30 s, 4°C) to ensure that all organoids are in the same z-plane and incubated overnight at 37°C. When the use of ECM domes is preferred, we recommend using low volume domes of 5 µL and low organoid density to avoid overlap of organoids in different z-planes. All drugs (Supplementary Table 2) and fluorescent reagents were added to the plate using the Tecan D300e Digital Dispenser and dissolved in DMSO. Cytotox Green (60 nM/well, Sartorius, #4633, DMSO) was used as a fluorescent cell death marker and Staurosporine (2 µM, Tocris Bioscience, #1285, DMSO) as a positive control. For each drug in the 3x3 synergy matrix, a 3-point logarithmic monotherapy titration was dispensed with the following concentration ranges: 100 nM, 547.7 nM, 3000 nM. For the 4x1 synergy matrix (PI3Ki screen), a 4-point logarithmic titration was used with the following concentrations: 100 nM, 311 nM, 965 nM and 3000nM and 1000 nM MRTX1133. DMSO concentrations were normalized to the same level in each well (< 1%). Brightfield and green fluorescence (Green channel: 461–487 nm/500–530 nm (excitation/emission)) wholewell images (4x objective) were taken at 0, 72 and 120 h with the Tecan Spark Cyto set at 37°C/5% CO2. Image and data analysis Images were analyzed with the Orbits ® label-free organoid detection module ( 43 ). The Normalized Organoid Growth Rate (NOGR) drug response metric was used for all downstream data analysis, which is described in detail in Deben et al., 2024 ( 45 ). Based on the NOGR, the drug effects can be classified as: >1, proliferative effect; = 1, normal growth as in negative control; = 0, complete growth inhibition; = -1, complete killing as in positive control. For synergy, a new derived variable Normalized (N)NOGR was computed to scale NOGR values between 0 and 100. The formula used for this computation is as follows: $$\:\frac{NNOGR}{NNDR}=\left(\frac{NOGR}{NDR}+1\right)*\:50$$ The ZIP/Bliss/Loewe/HSA synergy scores were calculated using the SynergyFinder R-package ( 46 ). A synergy score > 10: Indicates a synergistic interaction between the drugs. -10 < Score < 10: Implies an additive effect where the combined impact of the drugs is approximately equal to their individual effects summed. <-10: Signifies an antagonistic interaction between the drugs. Bulk RNA sequencing For RNA sequencing (RNA-seq), full grown organoids organoids were harvested after 5 days of culture in ECM domes. Afterwards, RNA was extracted using RNeasy midi kit (Qiagen). For removal of gDNA, RNAse-free DNAse treatment was performed. RNA concentration and purity were checked using the Qubit RNA BR Assay Kit on Qubit 4 Fluorometer (ThermoFisher) and NanoDrop ND-1000 (ThermoFisher), respectively. Samples were frozen at -80°C and delivered to Genomics Core Leuven for transcriptome sequencing using Lexogen QuantSeq 3’ FWD library preparation kit for Illumina on a Hiseq400 SR50 line with a minimum of 2M reads per sample. Downstream analysis and plotting were performed using the Omics Playground tool for feature-level clustering (Big Omics Analytics). Single cell RNA sequencing Single-cell RNA sequencing (scRNA-seq) data for the PDAC dataset were obtained from the Gene Expression Omnibus under accession number GSE205013 ( 14 ). Data preprocessing and analysis were conducted using the Python-based Scanpy framework. Initially, quality control filtering was applied to retain cells with at least 200 detected genes and to exclude genes expressed in fewer than 3 cells. To remove low-quality cells, we identified and excluded mitochondrial, ribosomal, and hemoglobin genes based on established gene name conventions. QC metrics such as total counts, number of genes, and mitochondrial gene content were calculated. Cells with high mitochondrial content (> 20%) were removed to ensure high-quality input data. After filtering, gene expression values were normalized to correct for differences in sequencing depth, followed by logarithmic transformation to stabilize variance across genes. The resulting expression matrix was scaled, and dimensionality reduction techniques including principal component analysis (PCA) and Uniform Manifold Approximation and Projection (UMAP) were applied to visualize the complex cellular heterogeneity within the dataset. Cancer cells were defined based on positive expression of canonical epithelial markers KRT19 and EPCAM, and these were retained for downstream analyses. To assess the activity of specific gene sets and biological pathways across individual cells, we employed AUCell algorithm, which computes the Area Under the Curve (AUC) for the cumulative expression of input gene sets per cell, allowing robust pathway activity estimation at single-cell resolution. The resulting AUC scores were visualized in low-dimensional space using UMAP embeddings, and further analyzed across clusters and patient subgroups. Functional enrichment and pathway analyses were conducted to identify distinct transcriptional programs and signaling pathway activations among cancer cell subpopulations. Abbreviations AOC Area over the curve AUC Area under the curve HSA Highest single agent NOGR Normalized organoid growth rate NSCLC Non-small cell lung cancer PCA principal component analysis PDAC Pancreatic ductal adenocarcinoma PDO Patient-derived organoid RNA-seq RNA sequencing scRNA-seq Single cell RNA sequencing UMAP Uniform manifold approximation and projection Declarations Ethics approval and consent to participate This study was approved by the Ethics Committee of the University Hospital Antwerp (UZA) under protocols 17/30/339 and 14/47/480. Written informed consent was obtained from all participating patients in accordance with institutional and national ethical guidelines. Consent for publication NA Availability of data and materials The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request. Competing interests The authors declare no conflict of interest. Funding This work is supported by: Research Foundation Flanders (FWO), Grant number; FWO-SB 1S27021N to M.L.C. and University Research Fund (BOF) of the University of Antwerp, Antwerp, Belgium, Grant number; FFB220225 to H.P. and S.S.. Part of the work was also supported by several donors, amongst which Willy Floren and Dedert Schilde vzw. The funders played no role in study design, data collection, analysis and interpretation of data, or the writing of this manuscript. Authors' contributions M.L.C. conceived the study. M.L.C., F.R.F. and S.S. performed the experiments and data acquisition. G.R., V.H., J.H., P.V.S., and N.K. provided patient samples and clinical data. M.L.C. and S.S. analyzed the data and wrote the manuscript under the supervision of C.D.. H.V., M.P., E.S., F.L., H.P. and C.D. supervised the project and provided critical feedback throughout the study. All authors reviewed and approved the final manuscript. Acknowledgements NA References Siegel RL, Kratzer TB, Giaquinto AN, Sung H, Jemal A. Cancer statistics, 2025. CA Cancer J Clin. 2025;75(1):10-45. Ramaswamy A, Srinivas S, Chaudhari V, Bhargava P, Bhandare M, Shrikhande SV, et al. Systemic therapy in pancreatic ductal adenocarcinomas (PDACs)-basis and current status. Ecancermedicalscience. 2022;16:1367. Zhang XP, Xu S, Gao YX, Zhao ZM, Zhao GD, Hu MG, et al. Early and late recurrence patterns of pancreatic ductal adenocarcinoma after pancreaticoduodenectomy: a multicenter study. Int J Surg. 2023;109(4):785-93. Integrated Genomic Characterization of Pancreatic Ductal Adenocarcinoma. Cancer Cell. 2017;32(2):185-203.e13. de Langen AJ, Johnson ML, Mazieres J, Dingemans AC, Mountzios G, Pless M, et al. Sotorasib versus docetaxel for previously treated non-small-cell lung cancer with KRAS(G12C) mutation: a randomised, open-label, phase 3 trial. Lancet. 2023;401(10378):733-46. Jänne PA, Riely GJ, Gadgeel SM, Heist RS, Ou SI, Pacheco JM, et al. Adagrasib in Non-Small-Cell Lung Cancer Harboring a KRAS(G12C) Mutation. N Engl J Med. 2022;387(2):120-31. Wei D, Wang L, Zuo X, Maitra A, Bresalier RS. A Small Molecule with Big Impact: MRTX1133 Targets the KRASG12D Mutation in Pancreatic Cancer. Clin Cancer Res. 2024;30(4):655-62. Isermann T, Sers C, Der CJ, Papke B. KRAS inhibitors: resistance drivers and combinatorial strategies. Trends in Cancer. 2025;11(2):91-116. Brown WS, McDonald PC, Nemirovsky O, Awrey S, Chafe SC, Schaeffer DF, et al. Overcoming Adaptive Resistance to KRAS and MEK Inhibitors by Co-targeting mTORC1/2 Complexes in Pancreatic Cancer. Cell Reports Medicine. 2020;1(8):100131. Liu X, Wang W, Liu X, Zhang Z, Yu L, Li R, et al. Multi-omics analysis of intra-tumoural and inter-tumoural heterogeneity in pancreatic ductal adenocarcinoma. Clinical and Translational Medicine. 2022;12(1):e670. Driehuis E, van Hoeck A, Moore K, Kolders S, Francies HE, Gulersonmez MC, et al. Pancreatic cancer organoids recapitulate disease and allow personalized drug screening. Proc Natl Acad Sci U S A. 2019;116(52):26580-90. Huang T, Cao H, Liu C, Sun X, Dai S, Liu L, et al. MAL2 reprograms lipid metabolism in intrahepatic cholangiocarcinoma via EGFR/SREBP-1 pathway based on single-cell RNA sequencing. Cell Death Dis. 2024;15(6):411. Wolf FA, Angerer P, Theis FJ. SCANPY: large-scale single-cell gene expression data analysis. Genome Biology. 2018;19(1):15. Werba G, Weissinger D, Kawaler EA, Zhao E, Kalfakakou D, Dhara S, et al. Single-cell RNA sequencing reveals the effects of chemotherapy on human pancreatic adenocarcinoma and its tumor microenvironment. Nat Commun. 2023;14(1):797. Wang X, Allen S, Blake JF, Bowcut V, Briere DM, Calinisan A, et al. Identification of MRTX1133, a Noncovalent, Potent, and Selective KRASG12D Inhibitor. Journal of Medicinal Chemistry. 2022;65(4):3123-33. Petrenko O, Kirillov V, D'Amico S, Reich NC. Intratumor heterogeneity in KRAS signaling shapes treatment resistance. iScience. 2025;28(2):111662. McDaid WJ, Wilson L, Adderley H, Martinez-Lopez A, Baker MJ, Searle J, et al. The PI3K-AKT-mTOR axis persists as a therapeutic dependency in KRASG12D-driven non-small cell lung cancer. Molecular Cancer. 2024;23(1):253. Ihle NT, Byers LA, Kim ES, Saintigny P, Lee JJ, Blumenschein GR, et al. Effect of KRAS oncogene substitutions on protein behavior: implications for signaling and clinical outcome. J Natl Cancer Inst. 2012;104(3):228-39. Hallin J, Bowcut V, Calinisan A, Briere DM, Hargis L, Engstrom LD, et al. Anti-tumor efficacy of a potent and selective non-covalent KRASG12D inhibitor. Nature Medicine. 2022;28(10):2171-82. Lee JE, Woo MG, Jung KH, Kang YW, Shin SM, Son MK, et al. Combination Therapy of the Active KRAS-Targeting Antibody inRas37 and a PI3K Inhibitor in Pancreatic Cancer. Biomol Ther (Seoul). 2022;30(3):274-83. Misale S, Fatherree JP, Cortez E, Li C, Bilton S, Timonina D, et al. KRAS G12C NSCLC Models Are Sensitive to Direct Targeting of KRAS in Combination with PI3K Inhibition. Clin Cancer Res. 2019;25(2):796-807. Suh KJ, Ryu MH, Zang DY, Bae WK, Lee HS, Oh HJ, et al. AZD8186 in Combination With Paclitaxel in Patients With Advanced Gastric Cancer: Results From a Phase Ib/II Study (KCSG ST18-20). Oncologist. 2023;28(9):e823-e34. Choudhury AD, Higano CS, de Bono JS, Cook N, Rathkopf DE, Wisinski KB, et al. A Phase I Study Investigating AZD8186, a Potent and Selective Inhibitor of PI3Kβ/δ, in Patients with Advanced Solid Tumors. Clin Cancer Res. 2022;28(11):2257-69. Marqués M, Tranchant R, Risa-Ebrí B, Suárez-Solís ML, Fernández LC, Carrillo-de-Santa-Pau E, et al. Combined MEK and PI3K/p110β Inhibition as a Novel Targeted Therapy for Malignant Mesothelioma Displaying Sarcomatoid Features. Cancer Res. 2020;80(4):843-56. Hwangbo H, Patterson SC, Dai A, Plana D, Palmer AC. Additivity predicts the efficacy of most approved combination therapies for advanced cancer. Nature Cancer. 2023;4(12):1693-704. Ning C, Liang M, Liu S, Wang G, Edwards H, Xia Y, et al. Targeting ERK enhances the cytotoxic effect of the novel PI3K and mTOR dual inhibitor VS-5584 in preclinical models of pancreatic cancer. Oncotarget. 2017;8(27):44295-311. Dolly SO, Wagner AJ, Bendell JC, Kindler HL, Krug LM, Seiwert TY, et al. Phase I Study of Apitolisib (GDC-0980), Dual Phosphatidylinositol-3-Kinase and Mammalian Target of Rapamycin Kinase Inhibitor, in Patients with Advanced Solid Tumors. Clin Cancer Res. 2016;22(12):2874-84. Powles T, Lackner MR, Oudard S, Escudier B, Ralph C, Brown JE, et al. Randomized Open-Label Phase II Trial of Apitolisib (GDC-0980), a Novel Inhibitor of the PI3K/Mammalian Target of Rapamycin Pathway, Versus Everolimus in Patients With Metastatic Renal Cell Carcinoma. J Clin Oncol. 2016;34(14):1660-8. Makker V, Recio FO, Ma L, Matulonis UA, Lauchle JO, Parmar H, et al. A multicenter, single-arm, open-label, phase 2 study of apitolisib (GDC-0980) for the treatment of recurrent or persistent endometrial carcinoma (MAGGIE study). Cancer. 2016;122(22):3519-28. Kim MJ, Lee SJ, Ryu JH, Kim SH, Kwon IC, Roberts TM. Combination of KRAS gene silencing and PI3K inhibition for ovarian cancer treatment. J Control Release. 2020;318:98-108. Yamamoto N, Fujiwara Y, Tamura K, Kondo S, Iwasa S, Tanabe Y, et al. Phase Ia/Ib study of the pan-class I PI3K inhibitor pictilisib (GDC-0941) administered as a single agent in Japanese patients with solid tumors and in combination in Japanese patients with non-squamous non-small cell lung cancer. Invest New Drugs. 2017;35(1):37-46. Sarker D, Ang JE, Baird R, Kristeleit R, Shah K, Moreno V, et al. First-in-human phase I study of pictilisib (GDC-0941), a potent pan-class I phosphatidylinositol-3-kinase (PI3K) inhibitor, in patients with advanced solid tumors. Clin Cancer Res. 2015;21(1):77-86. Shapiro GI, LoRusso P, Kwak E, Pandya S, Rudin CM, Kurkjian C, et al. Phase Ib study of the MEK inhibitor cobimetinib (GDC-0973) in combination with the PI3K inhibitor pictilisib (GDC-0941) in patients with advanced solid tumors. Invest New Drugs. 2020;38(2):419-32. Schöffski P, Cresta S, Mayer IA, Wildiers H, Damian S, Gendreau S, et al. A phase Ib study of pictilisib (GDC-0941) in combination with paclitaxel, with and without bevacizumab or trastuzumab, and with letrozole in advanced breast cancer. Breast Cancer Res. 2018;20(1):109. Soria JC, Adjei AA, Bahleda R, Besse B, Ferte C, Planchard D, et al. A phase IB dose-escalation study of the safety and pharmacokinetics of pictilisib in combination with either paclitaxel and carboplatin (with or without bevacizumab) or pemetrexed and cisplatin (with or without bevacizumab) in patients with advanced non-small cell lung cancer. Eur J Cancer. 2017;86:186-96. Krop IE, Mayer IA, Ganju V, Dickler M, Johnston S, Morales S, et al. Pictilisib for oestrogen receptor-positive, aromatase inhibitor-resistant, advanced or metastatic breast cancer (FERGI): a randomised, double-blind, placebo-controlled, phase 2 trial. Lancet Oncol. 2016;17(6):811-21. Leong S, Moss RA, Bowles DW, Ware JA, Zhou J, Spoerke JM, et al. A Phase I Dose-Escalation Study of the Safety and Pharmacokinetics of Pictilisib in Combination with Erlotinib in Patients with Advanced Solid Tumors. Oncologist. 2017;22(12):1491-9. Dilly J, Hoffman MT, Abbassi L, Li Z, Paradiso F, Parent BD, et al. Mechanisms of Resistance to Oncogenic KRAS Inhibition in Pancreatic Cancer. Cancer Discov. 2024;14(11):2135-61. André F, Ciruelos EM, Juric D, Loibl S, Campone M, Mayer IA, et al. Alpelisib plus fulvestrant for PIK3CA-mutated, hormone receptor-positive, human epidermal growth factor receptor-2-negative advanced breast cancer: final overall survival results from SOLAR-1. Ann Oncol. 2021;32(2):208-17. Jhaveri KL, Im S-A, Saura C, Loibl S, Kalinsky K, Schmid P, et al. Overall Survival with Inavolisib in PIK3CA-Mutated Advanced Breast Cancer. New England Journal of Medicine. 2025;393(2):151-61. Dienstmann R, Rodon J, Serra V, Tabernero J. Picking the Point of Inhibition: A Comparative Review of PI3K/AKT/mTOR Pathway Inhibitors. Molecular Cancer Therapeutics. 2014;13(5):1021-31. Le Compte M, De La Hoz EC, Peeters S, Fortes FR, Hermans C, Domen A, et al. Single-organoid analysis reveals clinically relevant treatment-resistant and invasive subclones in pancreatic cancer. NPJ Precis Oncol. 2023;7(1):128. Deben C, De La Hoz EC, Compte ML, Van Schil P, Hendriks JMH, Lauwers P, et al. OrBITS: label-free and time-lapse monitoring of patient derived organoids for advanced drug screening. Cell Oncol (Dordr). 2023;46(2):299-314. Le Compte M, Cardenas De La Hoz E, Peeters S, Smits E, Lardon F, Roeyen G, et al. Multiparametric Tumor Organoid Drug Screening Using Widefield Live-Cell Imaging for Bulk and Single-Organoid Analysis: 1940-087X; 2022. e64434 p. Deben C, Cardenas De La Hoz E, Rodrigues Fortes F, Le Compte M, Seghers S, Vanlanduit S, et al. Development and validation of the Normalized Organoid Growth Rate (NOGR) metric in brightfield imaging-based assays. Communications Biology. 2024;7(1):1612. Zheng S, Wang W, Aldahdooh J, Malyutina A, Shadbahr T, Tanoli Z, et al. SynergyFinder plus: toward better interpretation and annotation of drug combination screening datasets. Genomics, Proteomics and Bioinformatics. 2022;20(3):587-96. Additional Declarations No competing interests reported. Supplementary Files SupplementaryTable1.xlsx SupplementaryTable2.xlsx SupplementaryTable2.xlsx SupplementaryTable1.xlsx SupplementaryFile1.pdf SupplementaryFile2.pdf SupplementaryFigures.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7614104","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":524411986,"identity":"d3bfb9c3-7fcc-4163-b4d5-169a877c1bf4","order_by":0,"name":"Sofie Seghers","email":"","orcid":"","institution":"University of Antwerp","correspondingAuthor":false,"prefix":"","firstName":"Sofie","middleName":"","lastName":"Seghers","suffix":""},{"id":524411987,"identity":"8125aaf3-190c-4705-9d4a-eb37414b2ff9","order_by":1,"name":"Maxim Le Compte","email":"","orcid":"","institution":"University of Antwerp","correspondingAuthor":false,"prefix":"","firstName":"Maxim","middleName":"Le","lastName":"Compte","suffix":""},{"id":524411989,"identity":"dd35eb60-7e90-4707-b213-dd6717dc50cd","order_by":2,"name":"Felicia Rodrigues Fortes","email":"","orcid":"","institution":"University of Antwerp","correspondingAuthor":false,"prefix":"","firstName":"Felicia","middleName":"Rodrigues","lastName":"Fortes","suffix":""},{"id":524411990,"identity":"d11d475b-5e11-40f9-864f-07209b66be7d","order_by":3,"name":"Geert Roeyen","email":"","orcid":"","institution":"Antwerp University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Geert","middleName":"","lastName":"Roeyen","suffix":""},{"id":524411991,"identity":"132454b9-b565-4ed0-aa1e-3a38853ac36f","order_by":4,"name":"Vera Hartman","email":"","orcid":"","institution":"Antwerp University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Vera","middleName":"","lastName":"Hartman","suffix":""},{"id":524411993,"identity":"1762e58f-dd95-41f0-93fa-695eb7939948","order_by":5,"name":"Jeroen M.H. Hendriks","email":"","orcid":"","institution":"Antwerp University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Jeroen","middleName":"M.H.","lastName":"Hendriks","suffix":""},{"id":524411996,"identity":"0ed9b003-18e9-48b5-94bb-c8c1ef587373","order_by":6,"name":"Paul Van Schil","email":"","orcid":"","institution":"Antwerp University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Paul","middleName":"Van","lastName":"Schil","suffix":""},{"id":524411998,"identity":"05b4927f-007e-4c90-8d68-7cfc1423986a","order_by":7,"name":"Niels Komen","email":"","orcid":"","institution":"Antwerp University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Niels","middleName":"","lastName":"Komen","suffix":""},{"id":524412000,"identity":"f31a9797-3a05-494c-8707-f54cbbdcd85d","order_by":8,"name":"Hendrik Vercammen","email":"","orcid":"","institution":"University of Antwerp","correspondingAuthor":false,"prefix":"","firstName":"Hendrik","middleName":"","lastName":"Vercammen","suffix":""},{"id":524412001,"identity":"d20eff4d-9cee-4324-982a-e1da75ba3d8e","order_by":9,"name":"Marc Peeters","email":"","orcid":"","institution":"University of Antwerp","correspondingAuthor":false,"prefix":"","firstName":"Marc","middleName":"","lastName":"Peeters","suffix":""},{"id":524412003,"identity":"2d5bbb5f-aeec-41eb-8613-acf36c3c89ab","order_by":10,"name":"Evelien Smits","email":"","orcid":"","institution":"University of Antwerp","correspondingAuthor":false,"prefix":"","firstName":"Evelien","middleName":"","lastName":"Smits","suffix":""},{"id":524412005,"identity":"b578c683-77d8-44b8-bcdb-189bfaf81341","order_by":11,"name":"Filip Lardon","email":"","orcid":"","institution":"University of Antwerp","correspondingAuthor":false,"prefix":"","firstName":"Filip","middleName":"","lastName":"Lardon","suffix":""},{"id":524412006,"identity":"bde23d4b-a824-4a0d-a54b-8c7b313a04d7","order_by":12,"name":"Hans Prenen","email":"","orcid":"","institution":"Antwerp University Hospital","correspondingAuthor":false,"prefix":"","firstName":"Hans","middleName":"","lastName":"Prenen","suffix":""},{"id":524412007,"identity":"65471bf5-a9b5-42a5-8718-85a466cf5e0d","order_by":13,"name":"Christophe Deben","email":"data:image/png;base64,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","orcid":"","institution":"University of Antwerp","correspondingAuthor":true,"prefix":"","firstName":"Christophe","middleName":"","lastName":"Deben","suffix":""}],"badges":[],"createdAt":"2025-09-14 17:53:10","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":true,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-7614104/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7614104/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":92861949,"identity":"f6509506-f326-455d-bcde-0d360de2a103","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":62544992,"visible":true,"origin":"","legend":"","description":"","filename":"250916ManuscriptMRTXnumbered.docx","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/05c5ddeac291c9150652d2d0.docx"},{"id":92861917,"identity":"3b49d33b-3476-42fc-a134-c6a2b62de374","added_by":"auto","created_at":"2025-10-06 12:27:57","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":22142,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable1d1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/d9a78021af0e6331a43f10f6.xlsx"},{"id":92861919,"identity":"f3b11ec6-5980-4ebe-9b9c-cb04bc083fa5","added_by":"auto","created_at":"2025-10-06 12:27:57","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":18605,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable2d1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/66ac48820384dff95dac56ce.xlsx"},{"id":92862244,"identity":"157913ad-4c21-4efa-851c-6b1cb0627c82","added_by":"auto","created_at":"2025-10-06 12:35:57","extension":"json","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13151,"visible":true,"origin":"","legend":"","description":"","filename":"0f8ba3080c504b598515d0974beb50d9.json","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/1998d21e87f64c5ba3eeb5b2.json"},{"id":92861924,"identity":"b6152c51-3a76-4149-b3b1-cd238633eb5b","added_by":"auto","created_at":"2025-10-06 12:27:57","extension":"pdf","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":605292,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFile1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/d4e46c364aae84f4fa83340b.pdf"},{"id":92861929,"identity":"75c634d8-a954-46f5-9c24-41b093a41fee","added_by":"auto","created_at":"2025-10-06 12:27:57","extension":"pdf","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":356833,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFile2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/b920e3fd16799fec8ff9719d.pdf"},{"id":92861931,"identity":"363ab5fb-c262-4219-bb63-1f7e6fa45847","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":22142,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/ca4341e2be3ab948324407dc.xlsx"},{"id":92861941,"identity":"74f02191-948a-4d82-b037-3bc60f545a6f","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":18605,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/548d5728932b7d3a1c9d9c1b.xlsx"},{"id":92862249,"identity":"610c375c-2006-4c58-8886-ee23d99e81a6","added_by":"auto","created_at":"2025-10-06 12:35:58","extension":"xml","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":140683,"visible":true,"origin":"","legend":"","description":"","filename":"0f8ba3080c504b598515d0974beb50d91enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/46c07daf5a5d3239589bda3f.xml"},{"id":92862246,"identity":"592de543-f002-406f-9b38-b6b4cc42a096","added_by":"auto","created_at":"2025-10-06 12:35:57","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4821232,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/9d2dc53cf4cce9048d1b517e.png"},{"id":92862253,"identity":"e5c74954-0e3b-4ecb-bd5b-e9380200c4dd","added_by":"auto","created_at":"2025-10-06 12:35:58","extension":"emf","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4589788,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.emf","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/440968bb93917277348ec013.emf"},{"id":92861944,"identity":"577a5004-d6e7-452e-b90f-99d8054e4e20","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"emf","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2271624,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.emf","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/75b1254c1bdc029e4ca72963.emf"},{"id":92862250,"identity":"1ca70c64-49ff-4685-a98c-a45c8f0254c8","added_by":"auto","created_at":"2025-10-06 12:35:58","extension":"emf","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1764668,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.emf","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/fc67c26f3357e7d15b1c22de.emf"},{"id":92861950,"identity":"68221bfd-63cc-49ff-ba80-af3c3e054d6b","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"emf","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":46549304,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.emf","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/88b5c4bd60efd6c4b9ec9dd3.emf"},{"id":92863877,"identity":"677cc6d8-c9cd-4b4f-8064-48244b6de8ce","added_by":"auto","created_at":"2025-10-06 12:43:57","extension":"jpeg","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1240549,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/e5e4f019c6ab5129eac1c030.jpeg"},{"id":92862256,"identity":"a915e955-d175-4225-a71c-7b533acd5ad1","added_by":"auto","created_at":"2025-10-06 12:35:58","extension":"emf","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3199868,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.emf","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/71ad1f1020fad2822c396564.emf"},{"id":92862255,"identity":"9a6510dc-6747-4bd7-906c-5a7cea52246a","added_by":"auto","created_at":"2025-10-06 12:35:58","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3833167,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/12a8023495dd7add79910444.png"},{"id":92861934,"identity":"53fbddbd-7802-4ec4-a18d-667a06d57acf","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":489687,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/bfd0b30e34a24a2cd030e2c0.png"},{"id":92861939,"identity":"1341927b-f78d-418f-bdcf-7c4204bb954f","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":34176,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/b68bbb71f3d287fb5dfc5176.png"},{"id":92861933,"identity":"b4caba84-4ad1-491e-b504-f6ce64955771","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12745,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/c8c2ccc20473abc34a6de499.png"},{"id":92863878,"identity":"6d62a6df-557f-4cde-870d-8d6be10f3744","added_by":"auto","created_at":"2025-10-06 12:43:58","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9915,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/2a4019a4bff34f4ea8b5495a.png"},{"id":92861927,"identity":"49e1d7aa-016d-4e93-81b1-e9302a31c973","added_by":"auto","created_at":"2025-10-06 12:27:57","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":24703,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/f4d2d77bd6a2cc96eb2d5eb0.png"},{"id":92861940,"identity":"ba41e490-f983-4681-8b53-21a2293e051b","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":296460,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/f50bc037c3c9a7cb79ecc873.png"},{"id":92861930,"identity":"dc09fc33-dd8d-461f-b8f7-712fb62e4b68","added_by":"auto","created_at":"2025-10-06 12:27:57","extension":"png","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42729,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/0fb4b94ef3dfa2c3dfc71f7d.png"},{"id":92862252,"identity":"278fd4ac-445f-4ca0-a6a0-e9b75d466379","added_by":"auto","created_at":"2025-10-06 12:35:58","extension":"png","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":414690,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/c65888906455422fc1e1ddb3.png"},{"id":92861937,"identity":"296d5ced-0b1e-45f6-96b9-82c0be853806","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"xml","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":139526,"visible":true,"origin":"","legend":"","description":"","filename":"0f8ba3080c504b598515d0974beb50d91structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/ed7a936567c425bcc4793491.xml"},{"id":92861942,"identity":"69d49699-4cf8-4f08-afb3-9d972f7cd4de","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"html","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":149745,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/068dc9bcebee2af67867ffa5.html"},{"id":93629405,"identity":"f2c12aae-4d3c-457a-b9b7-0d3cc38b86b8","added_by":"auto","created_at":"2025-10-15 21:01:31","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":774801,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/3bb8c64a-d8ce-4381-aa65-0b86bd634b94.pdf"},{"id":92861916,"identity":"3f8c7f61-7421-49d6-b697-b4adc21113fa","added_by":"auto","created_at":"2025-10-06 12:27:57","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":22142,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/27e7d5970b5e4dd39ae9ac71.xlsx"},{"id":92861918,"identity":"fbf6567d-336e-44e5-b7a4-1e9f5af0c944","added_by":"auto","created_at":"2025-10-06 12:27:57","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":18605,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/9d2050addb73875771a63aef.xlsx"},{"id":92861922,"identity":"556def00-8017-48ef-85b5-b54894f38532","added_by":"auto","created_at":"2025-10-06 12:27:57","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":18605,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/8739a44847230910731de53e.xlsx"},{"id":92862245,"identity":"6a7aac26-1e52-4c32-ba15-84dc4a63b2c0","added_by":"auto","created_at":"2025-10-06 12:35:57","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":22142,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/65984ae7e6a59c4696de449f.xlsx"},{"id":92861923,"identity":"a93b7d96-4ca8-407f-ad2e-424cf2abd450","added_by":"auto","created_at":"2025-10-06 12:27:57","extension":"pdf","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":605292,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFile1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/018953ba9831fdf1475b9d3c.pdf"},{"id":92862247,"identity":"fa420ed8-aff0-4b8f-8f1b-c2c94f9c2b55","added_by":"auto","created_at":"2025-10-06 12:35:57","extension":"pdf","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":356833,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFile2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/c350cbf0b86482436dca7ea7.pdf"},{"id":92861946,"identity":"a21a65cb-bace-4335-9714-12589dea655f","added_by":"auto","created_at":"2025-10-06 12:27:58","extension":"docx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":7303318,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-7614104/v1/b49beee1b052c1af8e365f26.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Pathway-selective strategies reveal specific PI3K–mTOR inhibitors as key partners in KRAS G12D-targeted therapy","fulltext":[{"header":"Background","content":"\u003cp\u003ePancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive and treatment-resistant solid tumors, with a five-year survival rate around 13% and limited improvement in outcomes over recent decades (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). The disease is typically diagnosed at an advanced stage and remains refractory to most systemic therapies (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). Even among the minority of patients eligible for surgical resection, recurrence is frequent and prognosis remains poor (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). Thus, there is an urgent need for more effective, personalized therapeutic strategies. Mutations in the KRAS oncogene occur in over 90% of PDAC cases (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e), making it the most ubiquitous and well-established driver of pancreatic tumorigenesis. Historically considered \u0026ldquo;undruggable,\u0026rdquo; recent advances have led to the development of small molecules that directly target specific KRAS mutants. Notably, inhibitors against KRAS G12C, such as sotorasib and adagrasib, have demonstrated clinical efficacy in non-small cell lung cancer (NSCLC) and other malignancies (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). However, these inhibitors are not broadly applicable to PDAC, where KRAS G12D is the most common mutation, present in approximately 40% of patients (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e). Emerging compounds such as MRTX1133, which specifically target KRAS G12D, have shown promise in preclinical studies (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). Yet, KRAS inhibition in PDAC often leads to disease progression and resistance in most patients (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). This resistance underscores a fundamental challenge in targeting KRAS: despite its central role in tumor initiation, PDAC cells frequently develop alternative or compensatory signaling dependencies that undermine the durability of single-agent therapies (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). One contributing factor to therapeutic resistance is the high degree of inter- and intra-tumoral heterogeneity observed in PDAC (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). To address this complexity, there is growing interest in functional precision medicine approaches that integrate real-time drug testing with molecular profiling to identify patient-specific vulnerabilities. Patient-derived organoids (PDOs) have emerged as powerful \u003cem\u003eex vivo\u003c/em\u003e models that retain the histological and genetic features of the original tumor while enabling high-throughput drug screening (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e). Combined with advances in single-cell RNA sequencing (scRNA-seq) and pathway activity analysis, PDO platforms offer a unique opportunity to capture both phenotypic and molecular heterogeneity and guide the rational design of combination therapies (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). In this study, we used a multimodal strategy to dissect the functional and molecular heterogeneity of KRAS-mutant PDAC and identify effective combination strategies to enhance the limited efficacy of KRAS G12D inhibition. We first evaluated the therapeutic response of patient-derived organoids to MRTX1133 and uncovered variable, predominantly cytostatic effects. We then explored a publicly available scRNA seq database of PDAC patient samples, to explore underlying signaling diversity and identify actionable pathway dependencies at single cell level. Finally, we designed and screened a panel of drug combinations, based on the pathway analysis from scRNAseq data, revealing patient-specific and compound-specific synergies.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eKRAS G12D inhibition shows limited cytotoxic efficacy in \u003cem\u003eex vivo\u003c/em\u003e models\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate the therapeutic potential of KRAS G12D inhibition in pancreatic cancer, we treated four patient-derived organoid models harboring KRAS G12D (n=3) or G12V (n=1) mutations with increasing concentrations of the KRAS G12D-specific inhibitor MRTX1133. We selected 120h as the optimal timepoint for quantifying the growth rate\u0026ndash;based drug response metric normalized organoid growth rate (NOGR), as the observed variability in organoid growth among vehicle controls justifies the use of growth rate\u0026ndash;corrected metrics over traditional relative viability measures (Figure 1A). The NOGR metric provides an accurate assessment of drug response by distinguishing between cytostatic effects (NOGR between 1 and 0) and cytotoxic effects (NOGR between 0 and -1). Even at higher doses (up to 3 mM), G12D-mutant organoids showed only partial growth inhibition without significant loss of viability, indicated by a positive NOGR value (Figure 1B). This cytostatic effect was also visually confirmed, as shown in Figure 1C. PDAC002 (KRAS G12V) remained largely unaffected, confirming inhibitor specificity. However, G12D organoids began to recover and resume proliferation over time, indicating only a transient growth-inhibitory effect, as exemplified by PDAC044 (Figure 1D). Live-cell imaging at 3 \u0026mu;M MRTX1133 further supported this observation, revealing a cytostatic rather than cytotoxic response and subsequent growth recovery in PDAC044 organoids (Figure 1E). Collectively, these findings indicate that KRAS G12D inhibition can transiently suppress tumor growth but has limited therapeutic efficacy as monotherapy, aligning with an increasing number of drug combination clinical trials in KRAS mutant patients (NCT07020221, NCT0694976, NCT05382559,\u0026hellip;).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRational combination therapy development through single-cell RNA-seq and pathway profiling\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGiven the predominantly cytostatic response to KRAS G12D inhibition observed in patient-derived organoids (Figure 1), we sought to uncover co-targetable vulnerabilities that might convert this effect into a cytotoxic response. Specifically, we aimed to address two key questions: \u0026nbsp;is KRAS pathway activity heterogeneously distributed across tumor cells within PDAC tumors and can single-cell transcriptomic data be used to rationally identify novel combination therapy strategies tailored to the diverse signaling dependencies present in these tumors. To this end, we explored a publicaly available scRNA-seq dataset of PDAC tumors using Scanpy (13, 14). After normalization and quality control, we annotated epithelial tumor cells based on canonical epithelial markers, including EPCAM and KRT19. UMAP visualization of the integrated datasets, each representing an individual patient (Figure 2A, left), revealed significant inter- and intra-patient heterogeneity, with epithelial cells scattered across multiple distinct clusters (Figure 2A, right), suggesting functional and phenotypic diversity within and between tumors. Differential expression analysis revealed cluster-specific marker genes (Figure 2B), underscoring transcriptional divergence among tumor cells, supporting the hypothesis that distinct subpopulations may underlie patient-specific co-targetable vulnerabilities.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo further evaluate the functional heterogeneity, we calculated global pathway activity scores for each patient using gene sets from Hallmark, Reactome, and WikiPathways (Figure 2C). This analysis showed clear differences in signaling pathway activation between patients, including variability in cell cycle control, stress response, and growth factor signaling. Next, we examined KRAS-specific signaling activity at single cell level using AUCell. As shown in Figure 2D, HALLMARK_KRAS_SIGNALING_UP scores varied widely across tumor clusters, confirming that KRAS-driven transcriptional programs are not uniformly active, even within tumors harboring KRAS mutations. In contrast, HALLMARK_KRAS_SIGNALING_DN scores were consistently low and uniformly distributed across all clusters, suggesting a broadly repressed transcriptional program that is typically downregulated when KRAS signaling is active. This pattern implies that while KRAS-driven oncogenic programs may differ in amplitude between subpopulations, a baseline level of KRAS activity is maintained across the tumor. Such a profile may reflect a minimal threshold of KRAS dependency common to all malignant epithelial cells, with superimposed heterogeneity arising from additional pathway activations. Similarly, transcriptional activity of tyrosine kinase pathways (REACTOME_SIGNALING_BY_RECEPTOR_TYROSINE_KINASES) also showed a rather \u0026lsquo;patchy\u0026rsquo; distribution, indicating that kinase signaling dependencies differ at the single-cell level. To capture these functional differences in a more integrated manner, we performed PCA on global pathway scores and identified five distinct tumor subtypes (Figure 2E). Clustering of these subtypes revealed that each was characterized by unique pathway activation profiles based on their top 7 activated pathways (Figure 2F), including differential enrichment of MYC targets, E2F targets, EGFR, FGFR, and MAPK/ERK signaling, all pathways with known roles in PDAC progression and resistance. Based on these findings, we rationally designed a targeted drug panel (Figure 2G) that reflects the dominant signaling axes active within the different tumor subpopulations. This strategy offers a path toward more effective and personalized combination therapies by directly addressing the observed functional heterogeneity. In summary, our single-cell transcriptomic analysis lays the groundwork for rationally designed combination treatments with KRAS inhibitors, tailored to the specific signaling dependencies of PDAC subtypes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eKinase inhibitor screening confirms functional heterogeneity predicted by single-cell transcriptomics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on the signaling pathway heterogeneity identified in our single-cell transcriptomic analysis (Figure 2D), we next examined whether this patient-specific heterogeneity was also captured in our organoid panel and whether it could be linked to differences in drug response. Geneset signature maps were generated using UMAP clustering of genes by covariance, with coloring based on relative (average) log-expression in the phenotype group (Figure S1, supplementary table 1). This projection highlights geneset co-expression patterns and underscores the marked heterogeneity in pathway activity across organoid lines. Figure 3A provides an overview of selected genesets relevant to the drug panel shown in Figure 2G, revealing enhanced enrichment of these pathways in PDAC060 compared to the other organoid lines. Notably, a similar sample-level clustering pattern was observed in the drug response data, where PDAC060 exhibited marked hypersensitivity to the majority of tested compounds, as reflected by the Area Over the Curve (AOC) of the NOGR dose-response curves (Figure 3B, Figure S2). The MEK inhibitors trametinib (selective) and ASTX029 (dual MEK/ERKi) showed the strongest anticancer activity across all organoid lines, with PDAC060 being the most sensitive line (Figure 3B). A detailed compound list is provided in Supplementary Table 2.\u003c/p\u003e\n\u003cp\u003eTo quantify the extent of response variability, we ranked all tested compounds by the standard deviation of their AOC values across organoid lines (Figure 3C). Compounds such as sitravatinib, and trametinib showed a great degree of variability, suggesting they may be particularly influenced by tumor-specific signaling contexts. Finally, a Pearsons correlation analysis comparing drug responses to MRTX1133 revealed both overlapping and distinct response patterns (Figure 3D), providing insight into potential combinatorial strategies or resistance mechanisms. Together, these data confirm that the signaling heterogeneity observed at the transcriptomic level is also observed in the differential drug sensitivities. This reinforces the rationale for personalized combination therapies targeting dominant pathways within each tumor subtype.\u003c/p\u003e\n\u003ch3\u003eSynergy screening identifies MAPK\u0026ndash;ERK and PI3K\u0026ndash;mTOR axis combinations as promising therapeutic strategies in PDAC\u003c/h3\u003e\n\u003cp\u003eTo build on the observed pathway heterogeneity, we screened rational drug combinations in four KRAS-mutant PDAC organoid models, focusing on inhibitors of MAPK, PI3K\u0026ndash;AKT\u0026ndash;mTOR, and related kinase pathways. This to evaluate whether dual targeting could overcome the cytostatic limitations observed with KRAS G12D monotherapy. To assess whether observed synergy was specifically dependent on KRAS G12D inhibition or could also arise through parallel pathway interactions, we included a subset of combinations using paxalisib (a dual PI3K\u0026ndash;mTOR inhibitor) as the backbone compound. This allowed us to decouple synergy from direct G12D targeting and evaluate whether convergent inhibition of survival signaling (e.g., PI3K\u0026ndash;mTOR + MAPK) could produce similar or stronger effects in models with alternative KRAS mutations, such as PDAC002 (KRAS G12V).\u003c/p\u003e\n\u003cp\u003eSynergy between drug pairs was evaluated using a 3x3 synergy matrix and quantified using four established reference models: Highest Single Agent (HSA), Loewe, Bliss, and ZIP. As shown in Figure 4A, we observed a high degree of correlation between the HSA and Loewe models (Pearson r \u0026gt; 0.95), suggesting strong agreement in synergy classification across these two metrics. Similarly, the ZIP and BLISS models correlated most with each other. Violin plots of synergy scores across all combinations (Figure 4B) revealed that higher synergy scores were obtained with HSA and Loewe, consistent with previous reports that ZIP and BLISS may be more conservative in detecting synergy. We selected HSA for further downstream analysis, focusing on combinations that outperform the most effective single agent, as it provides a more interpretable and less conservative estimate of synergy compared to ZIP and BLISS models.\u003c/p\u003e\n\u003cp\u003eAll individual synergy matrices for each unique drug-organoid combination are provided in Supplementary File 1. Figure 4C summarizes the mean HSA synergy scores for drug combinations with 3 \u0026mu;M MRTX1133. The bubble plot visualizes both the level of synergy (bubble size and category) and the corresponding NOGR values (color gradient), allowing distinction between cytostatic (blue, large bubble) and cytotoxic (red, large bubble) synergistic interactions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eConsistent with the monotherapy responses, we observed organoid-specific synergistic drug combinations. As expected, no synergistic interactions with MRTX1133 were detected in PDAC002, which lacks the KRAS G12D mutation. In contrast, PDAC060, characterized by the strongest pathway activation among all organoid lines, exhibited the highest number of synergistic interactions across nearly all drug classes. Synergy in PDAC044 was particularly enriched for PI3K pathway-targeting drugs. PDAC087, on the other hand, was the most resistant, showing only limited cytotoxic synergy.\u003c/p\u003e\n\u003cp\u003eFor drug combinations involving the dual PI3K\u0026ndash;mTOR inhibitor paxalisib, synergy was primarily observed with MEK inhibitors such as trametinib and ASTX029. Notably, this included organoid lines like PDAC087, for which few synergistic combinations with MRTX1133 were identified, highlighting a potential alternative strategy. Overall, drug combinations targeting the PI3K/mTOR axis had the broadest activity in both PDAC044 and PDAC060. Therefore we selected this therapeutic class to expand the targeted drug panel.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSynergy with MRTX1133 is driven by chemical structure differences and not solely by target class\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBuilding on earlier observations that the PI3K/mTOR pathway represents a promising axis for combination therapy in KRAS G12D-mutant PDAC, we next aimed to dissect the compound-specific and contextual factors that shape synergy with MRTX1133. To explore this, we assembled a chemically diverse panel of 34 inhibitors targeting the PI3K/mTOR axis and tested these in a 4x1 synergy matrix (1 mM MRTX1133). This included isoform-specific PI3K inhibitors (e.g. taselisib, inavolisib), dual PI3K/mTOR inhibitors (e.g. omipalisib, gedatolisib) and pan-PI3K inhibitors (e.g. Pictilisib, AMG511). A detailed compound list is provided in Supplementary Table 2. All compounds were tested in combination with MRTX1133 across six cell models, including KRAS G12D- and G12V-mutant PDAC organoids, a KRAS G13D cholangiocarcinoma model (CCA012), and non-malignant lung organoids (LU046). This experimental setup enabled the prioritization of drug combinations that selectively target KRAS G12D mutant cancer cells while sparing healthy epithelial cells.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFigure 5A shows a clustered heatmap of monotherapy drug responses across organoid lines, based on the AOC of NOGR dose\u0026ndash;response curves (Supplemental Figure 3). Hierarchical clustering revealed distinct response patterns, with the most potent compounds clustering at the top, predominantly consisting of dual PI3K/mTOR inhibitors. Notably, healthy epithelial lung organoids (LU046) clustered closely with the most sensitive cancer organoid lines PDAC060 and CCA012, suggesting that these therapeutic classes lack selectivity for cancer cells in this conventional organoid model system. Figure 5B summarizes the mean HSA synergy scores for drug combinations with 1 \u0026mu;M MRTX1133 and the corresponding NOGR values. Detailed synergy matrices are provided in Supplementary File 2. \u0026nbsp;Synergistic effects were specifically observed in KRAS G12D mutant organoid lines, but were absent in KRAS G13D (CCA012), KRAS G12V (PDAC002), and healthy lung organoids (LU046). This selective synergy was evident with both PI3K inhibitors and dual PI3K/mTOR inhibitors, supporting the notion that effective combinations with MRTX1133 can be tumor-selective, and that synergy is not the result of additive pathway inhibition, but a result of tumor-specific dependencies and drug-structure interactions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo identify tumor selective drug combinations, we filtered for inhibitors with an NOGRmax \u0026gt; 0.5 (effect a 3\u0026nbsp;mM) in LU046, indicating a marginal growth inhibitory effect in healthy epithelial cells. Of the 10 inhibitors that met these criteria (Supplemental Figure 4A), only AZD8186 showed selective synergy: cytostatic in PDAC044, cytotoxic in PDAC060, and additive cytostatic in PDAC087 (Figure 5B (green), Figure 5C-D). As monotherapy, all organoid lines exhibited a predominantly cytostatic response profile (Figure 5E).\u003c/p\u003e\n\u003cp\u003eNotably, only apitolisib, pictilisib, and VS-5584 induced substantial synergistic effects across all three KRAS G12D-mutant organoid lines (Figure 5B (red), Figure S4B\u0026ndash;D). Among these, VS-5584 was the most potent, inducing a strong synergistic response at 300 nM in combination with 1 \u0026mu;M MRTX1133, as illustrated by the live-cell images (Figure S4E). However, VS-5584 also showed high monotherapy potency across all organoid lines, including the healthy epithelial lung organoids LU046 (Figure S4F), raising concerns about selectivity.\u003c/p\u003e\n\u003cp\u003eOf the four FDA-approved PI3K inhibitors tested (alpelisib, idelalisib, duvalisib and inavolisib), only alpelisib and inavolisib demonstrated synergistic interactions with MRTX1133 (Figure 5B, blue). Inavolisib was the most potent, showing consistent synergy across all tested concentrations in PDAC044 and PDAC060 (Figure S4G-H). Notably, inavolisib induced only a cytostatic effect as monotherapy in all organoid lines, including the healthy epithelial lung organoids LU046N.\u003c/p\u003e\n\u003cp\u003eFigure 5B highlights the compound- and model-specific nature of these responses. While omipalisib, voxtalisib, and PQR530 are all dual PI3K\u0026ndash;mTOR inhibitors, only omipalisib and PQR530 elicited synergy in specific models, whereas voxtalisib showed negligible activity, reinforcing that nominal target overlap does not predict functional synergistic outcome. In summary, these data reveal that monotherapy efficacy and synergy with KRAS G12D inhibition is not solely determined by PI3K\u0026ndash;mTOR target class, but rather by a combination of tumor-intrinsic signaling dependencies and compound-specific properties. Even among agents with similar reported targets, response profiles diverge significantly between models, reinforcing the need for chemotype-informed combination strategies that account for intertumoral heterogeneity and potential off-target effects of TKIs.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eRecent advances in the development of KRAS G12D inhibitors have re-energized efforts to directly target mutant KRAS in PDAC. MRTX1133 represented the first clinical inhibitor of KRAS G12D, with preclinical studies demonstrating significant tumor regression in KRAS G12D\u0026ndash;mutant models (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e). However, early clinical evaluation revealed modest and transient responses in patients, compounded by suboptimal pharmacokinetic properties, ultimately halting further development. These limitations reflect not only the pharmacological challenges inherent to targeting KRAS G12D but also the plasticity and heterogeneity of tumor signaling networks.\u003c/p\u003e\u003cp\u003eKRAS dependency in PDAC is neither binary nor uniform. Even within KRAS-mutant tumors, we observe substantial heterogeneity in reliance on canonical KRAS-driven pathways (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e). Single-cell RNA sequencing analysis of KRAS G12D\u0026ndash;mutant epithelial cells revealed uneven KRAS signaling activity, with distinct subpopulations exhibiting activation of alternative programs, including receptor tyrosine kinase signaling, MYC-driven transcriptional targets, and cellular stress response pathways. This intratumoral heterogeneity enables dynamic rewiring of signaling dependencies, facilitating adaptive resistance to KRAS G12D inhibition. Consistent with this, MRTX1133 monotherapy elicited predominantly cytostatic responses in vivo, likely reflecting the engagement of compensatory survival pathways upon KRAS inhibition.\u003c/p\u003e\u003cp\u003eThese findings underscore a fundamental limitation of single-agent targeted therapies in PDAC: inhibition of a single oncogenic node within a highly adaptive signaling network is insufficient to achieve durable tumor control. KRAS activating mutations drive strong dependency on the PI3K/AKT pathway (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e). Ihle et al. showed that KRAS G12D mutateted NSCLC models displayed elevated PI3K and MEK/ERK signaling (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e), and pharmacological and CRISPR-based screens identified PI3Kα as a critical bypass pathway whose co-inhibition with KRAS G12D enhances anti-tumor activity (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e). Consistently, work with the inRas37 antibody combined with the PI3K/mTOR inhibitor BEZ-235 in pancreatic cancer showed that dual inhibition of RAS/MAPK and PI3K pathways effectively blocked MAPK reactivation, significantly suppressing tumor growth in vitro and in vivo (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). To address this, we implemented rational combination strategies guided by functional drug screening and pathway activity inferred from single-cell transcriptomic data. Although patient-specific synergistic combinations were observed, such as MRTX1133 and ASTX029 in PDAC060, the most consistent synergy across patients was seen when MRTX1133 was combined with PI3K\u0026ndash;mTOR pathway inhibition. These findings are consistent with a recent study by Misale and colleagues, which demonstrated that among all targets screened, KRAS (G12C) blockade in combination with PI3K pathway inhibition in NSCLC produced the most robust therapeutic effect (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAmong the PI3k-mTOR combination regimen tested, AZD8186 (a PI3Kb/d inhibitor) showed the most tumor-selective synergy, but remained largely cytostatic, inducing growth arrest rather than cell death. This highlights an important consideration in combination therapy design: tumor-selective combinations can mitigate off-target toxicities and improve tolerability, but their clinical benefit may be limited if the synergy does not translate into meaningful cytotoxicity. This challenge is illustred by the halted phase Ib/II trial of AZD8186 in combination in advanced gastric cancer due to futility (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e). Nevertheless, a phase I trial combining AZD8186 with abiraterone or the mTOR inhibitor vistusertib reported preliminary anti-tumor activity (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e). Preclinical studies in Pten;Trp53-null mice further demonstrated that AZD8186 combined with selumetinib, a MEK inhibitor, improved survival and reduced mesothelioma proliferation (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e), supporting vertical co-targeting of the KRAS\u0026ndash;MEK axis and PI3K signaling to overcome resistance. In contrast to the tumor selective AZD8186, apitolisib (PI3K/mTORi), pictilisib (a pan-PI3Ki) and VS5584 (PI3K/mTOR), emerged as a highly synergistic partners that, despite lower tumor selectivity, induced more pronounced tumor cell death in combination. These findings emphasize that synergy alone does not guarantee clinical success. Even statistically significant interactions may fail to achieve the absolute level of tumor control required for durable responses if neither agent demonstrates sufficient monotherapy efficacy (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e). Of these three compounds, VS5584 was the most potent in KRAS G12D-mutant cells. While two clinical trials (NCT01991938, NCT02372227) were terminated early due to limited recruitment, preclinical data support its rational use. Notably, Ning et al. showed superior tumor control with VS-5584 plus an ERK inhibitor, underscoring the value of dual PI3K/mTOR and MAPK/ERK pathway blockade (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e). These findings underscore that synergy alone is insufficient unless paired with meaningful cytotoxic efficacy and mechanistic plausibility, especially in settings where vertical pathway inhibition may be needed to overcome adaptive resistance. Apitolisib has completed 12 clinical phase I/II trials, showing modest but durable activity in solid tumors (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e), though a trial in RCC found it less effective than everolimus (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e). Across studies, severe on-target toxicities (mainly hyperglycemia) have limited its clinical potential (\u003cspan additionalcitationids=\"CR28\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e). Pictilisib also shows promise: preclinical studies demonstrated synergy with siKRAS in KRAS G12D/PTEN-deficient models (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e), but clinical trials, both as monotherapy (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e) and in combination (\u003cspan additionalcitationids=\"CR34\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e), have reported mixed results (\u003cspan additionalcitationids=\"CR32 CR33 CR34 CR35 CR36\" citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e). While some trials reported acceptable tolerability with encouraging anti-tumor activity, others highlighted limited efficacy, often due to dose-limiting toxicities. Unlike the investigational PI3K inhibitors described above, alpelisib and inavolisib are clinically validated PI3Kα-selective inhibitors, offering a clear translational advantage for combination with KRAS G12D inhibitors. Both target PI3Kα-driven signaling, a key bypass mechanism upon KRAS inhibition (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e). Alpelisib, FDA-approved for HR+/HER2\u0026ndash; breast cancer with PIK3CA mutations (SOLAR-1 trial) (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e), effectively suppresses PI3K-mediated feedback activation of RTK/MAPK pathways, supporting its rationale in dual KRAS\u0026ndash;PI3K targeting. Inavolisib, a next-generation PI3Kα inhibitor with mutant PIK3CA degradation activity, has shown superior and durable pathway suppression, with recent INAVO120 data confirming clinical benefit in HR+/PIK3CA-mutant breast cancer (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e). Both agents showed strong cytotoxicity in 2 of 3 KRAS G12D-mutant PDAC lines when combined with MRTX1133, contrasting with the cytostatic effects of AZD8186. Their FDA approval and advanced clinical data lower translational barriers for combination trials, though toxicity management remains critical.\u003c/p\u003e\u003cp\u003eWhen comparing these classes of inhibitors, a key observation is the differential toxicity and activity of PI3K-targeting agents. Dual PI3K/mTOR inhibitors (e.g., gedatolisib, apitolisib, VS-5584) show the broadest activity, targeting multiple pathway nodes. However, this broad activity comes at the cost of greater toxicity, as also seen in our lung organoid models, making them less suitable for combination strategies and better suited for select single-agent use (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e). Pan-PI3K inhibitors have a narrower activity range but still present systemic toxicities (\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e), whereas isoform-specific inhibitors (e.g., alpelisib, inavolisib) offer a more favorable safety profile and could be suited for combination therapy.\u003c/p\u003e\u003cp\u003eFuture work should focus on preclinical validation of these combinations in vivo, exploration of predictive biomarkers, and ultimately, clinical testing to determine whether KRAS\u0026ndash;PI3K co-targeting can achieve durable responses in PDAC.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn summary, our study highlights the potential of combining KRAS G12D inhibition with PI3K pathway blockade to overcome the cytostatic limitations of MRTX1133 monotherapy. While dual PI3K/mTOR inhibitors show broad activity, their toxicity profile limits clinical translation, particularly in combination regimens. Conversely, isoform-specific inhibitors such as alpelisib and inavolisib not only demonstrated robust synergy with MRTX1133 in KRAS G12D-mutant PDAC models but also benefit from established clinical use and manageable safety profiles. These findings underscore the importance of rational inhibitor selection based on both tumor biology and tolerability.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003ePatient Material\u003c/h2\u003e\u003cp\u003eThe patient-derived organoids used in this study are registered in the Biobank@UZA (Antwerp, Belgium; ID: BE71030031000); Belgian Virtual Tumorbank funded by the National Cancer Plan. Organoids were derived from resection fragments or biopsies obtained from cancer patients treated at the Antwerp University Hospital and a written informed consent was obtained from all patients. The study was approved by the UZA Ethical Committee (ref. 17/30/339 and 14/47/480).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eOrganoid cultures\u003c/h2\u003e\u003cp\u003eBasic medium consisted of Ad-DF + ++ (Advanced DMEM/F12 (GIBCO, #12634028), with 1% GlutaMAX (GIBCO, 35050-061), 1% HEPES (GIBCO, #15630080), 1% penicillin/streptomycin (GIBCO, #15140122) supplemented with 2% Primocin (Invivogen, #ant-pm-05). For PDAC and CCA organoids, Ad-DF + ++ was supplemented with 0.5 nM WNT Surrogate-Fc-Fusion protein (ImmunoPrecise, #N001), 4% Noggin-Fc Fusion Protein conditioned medium (ImmunoPrecise, #N002), 4% Rspo3-Fc Fusion Protein conditioned medium (ImmunoPrecise, #R001), 1x B27 (Gibco, #17504044), 10 mM nicotinamide (Sigma-Aldrich, #N0636), 1.25 mM N-acetylcysteine (Sigma-Aldrich, #A9165), 100 ng/ml FGF-10 (Peprotech, #100\u0026thinsp;\u0026minus;\u0026thinsp;26), 500 nM A83-01 (Tocris, #2939), 10 nM gastrin (R\u0026amp;D Systems, #G9145) and 10 \u0026micro;M Y-27632 after passaging (Selleck Chemicals, #S1049). Normal pulmonary organoids were cultured in Ad-DF+++ supplemented with 4% Noggin-Fc Fusion Protein conditioned medium (ImmunoPrecise), 4% Rspo3-Fc Fusion Protein conditioned medium (ImmunoPrecise), 1x B27 (Gibco), 10 mM nicotinamide (Sigma-Aldrich), 1.25 mM N-acetylcysteine (Sigma-Aldrich), 100 ng/ml FGF-10 (Peprotech), 25 ng/ml FGF-7 (Peprotech), 500 nM A83-01 (Tocris) and 1 \u0026micro;M SB202190 (Sanbio, Cayman Chemical). For passaging, the organoids were digested to single cells with TrypLE Express (GIBCO, #12604021) and resuspended in \u0026gt;\u0026thinsp;80% ice cold Cultrex growth factor reduced BME type 2 (R\u0026amp;D Systems, #3533-010-02) in full organoid medium. Small droplets of 20 \u0026micro;L were plated and incubated inverted for 30 min at 37\u0026deg;C to allow them to solidify after which the drops were covered with full organoid medium. In-depth characterization of PDAC organoids used in this study is described in Le Compte et al., 2023 (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e) and the normal pulmonary organoid LU046 in Deben et al., 2023 (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eDrug screening\u003c/h2\u003e\u003cp\u003eDrug screening on 3D organoids was performed at the DrugVision.AI automated screening facility of the University of Antwerp, Belgium, using a prevalidated drug screening pipeline for which a detailed protocol is available in the Journal of Visualized Experiments (\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e). Briefly, established organoid lines were expanded in ECM domes (Cultrex type 2, R\u0026amp;D Systems, #3533-010-02). Next, 3-day-old organoids were harvested from ECM drops using the Cultrex Organoid Harvesting Solution (R\u0026amp;D Systems, #3700-100-01), collected in a 15 mL tube coated with 0.1% BSA/PBS, washed, and resuspended in medium. Next, the number of organoids was quantified by adding 5 \u0026micro;L of the organoid solution to 45 \u0026micro;L of medium in a 384-microplate well. A whole-well brightfield image was captured using the Tecan Spark Cyto and the number of organoids was counted label-free using Orbits\u003csup\u003e\u0026reg;\u003c/sup\u003e (Orbits Oncology). Next, the organoid solution was diluted in full medium supplemented with 4% Cultrex at a concentration of 4000 organoids / mL. Next, 50 \u0026micro;L (200 organoids) of this solution was dispensed into each well of a 384-well ultra-low attachment microplate (Corning, #4588) using the OT-2 pipetting robot (Opentrons) in a cooled environment. Thereafter, the plate was centrifuged (100 rcf, 30 s, 4\u0026deg;C) to ensure that all organoids are in the same z-plane and incubated overnight at 37\u0026deg;C. When the use of ECM domes is preferred, we recommend using low volume domes of 5 \u0026micro;L and low organoid density to avoid overlap of organoids in different z-planes. All drugs (Supplementary Table\u0026nbsp;2) and fluorescent reagents were added to the plate using the Tecan D300e Digital Dispenser and dissolved in DMSO. Cytotox Green (60 nM/well, Sartorius, #4633, DMSO) was used as a fluorescent cell death marker and Staurosporine (2 \u0026micro;M, Tocris Bioscience, #1285, DMSO) as a positive control. For each drug in the 3x3 synergy matrix, a 3-point logarithmic monotherapy titration was dispensed with the following concentration ranges: 100 nM, 547.7 nM, 3000 nM. For the 4x1 synergy matrix (PI3Ki screen), a 4-point logarithmic titration was used with the following concentrations: 100 nM, 311 nM, 965 nM and 3000nM and 1000 nM MRTX1133. DMSO concentrations were normalized to the same level in each well (\u0026lt;\u0026thinsp;1%). Brightfield and green fluorescence (Green channel: 461\u0026ndash;487 nm/500\u0026ndash;530 nm (excitation/emission)) wholewell images (4x objective) were taken at 0, 72 and 120 h with the Tecan Spark Cyto set at 37\u0026deg;C/5% CO2.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003eImage and data analysis\u003c/h2\u003e\u003cp\u003eImages were analyzed with the Orbits\u003csup\u003e\u0026reg;\u003c/sup\u003e label-free organoid detection module (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e). The Normalized Organoid Growth Rate (NOGR) drug response metric was used for all downstream data analysis, which is described in detail in Deben et al., 2024 (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e). Based on the NOGR, the drug effects can be classified as: \u0026gt;1, proliferative effect; = 1, normal growth as in negative control; = 0, complete growth inhibition; = -1, complete killing as in positive control.\u003c/p\u003e\u003cp\u003eFor synergy, a new derived variable Normalized (N)NOGR was computed to scale NOGR values between 0 and 100. The formula used for this computation is as follows:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\frac{NNOGR}{NNDR}=\\left(\\frac{NOGR}{NDR}+1\\right)*\\:50$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe ZIP/Bliss/Loewe/HSA synergy scores were calculated using the SynergyFinder R-package (\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e). A synergy score\u0026thinsp;\u0026gt;\u0026thinsp;10: Indicates a synergistic interaction between the drugs. -10\u0026thinsp;\u0026lt;\u0026thinsp;Score\u0026thinsp;\u0026lt;\u0026thinsp;10: Implies an additive effect where the combined impact of the drugs is approximately equal to their individual effects summed. \u0026lt;-10: Signifies an antagonistic interaction between the drugs.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003eBulk RNA sequencing\u003c/h2\u003e\u003cp\u003eFor RNA sequencing (RNA-seq), full grown organoids organoids were harvested after 5 days of culture in ECM domes. Afterwards, RNA was extracted using RNeasy midi kit (Qiagen). For removal of gDNA, RNAse-free DNAse treatment was performed. RNA concentration and purity were checked using the Qubit RNA BR Assay Kit on Qubit 4 Fluorometer (ThermoFisher) and NanoDrop ND-1000 (ThermoFisher), respectively. Samples were frozen at -80\u0026deg;C and delivered to Genomics Core Leuven for transcriptome sequencing using Lexogen QuantSeq 3\u0026rsquo; FWD library preparation kit for Illumina on a Hiseq400 SR50 line with a minimum of 2M reads per sample. Downstream analysis and plotting were performed using the Omics Playground tool for feature-level clustering (Big Omics Analytics).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003eSingle cell RNA sequencing\u003c/h2\u003e\u003cp\u003eSingle-cell RNA sequencing (scRNA-seq) data for the PDAC dataset were obtained from the Gene Expression Omnibus under accession number GSE205013 (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e). Data preprocessing and analysis were conducted using the Python-based Scanpy framework. Initially, quality control filtering was applied to retain cells with at least 200 detected genes and to exclude genes expressed in fewer than 3 cells. To remove low-quality cells, we identified and excluded mitochondrial, ribosomal, and hemoglobin genes based on established gene name conventions. QC metrics such as total counts, number of genes, and mitochondrial gene content were calculated. Cells with high mitochondrial content (\u0026gt;\u0026thinsp;20%) were removed to ensure high-quality input data. After filtering, gene expression values were normalized to correct for differences in sequencing depth, followed by logarithmic transformation to stabilize variance across genes. The resulting expression matrix was scaled, and dimensionality reduction techniques including principal component analysis (PCA) and Uniform Manifold Approximation and Projection (UMAP) were applied to visualize the complex cellular heterogeneity within the dataset. Cancer cells were defined based on positive expression of canonical epithelial markers KRT19 and EPCAM, and these were retained for downstream analyses. To assess the activity of specific gene sets and biological pathways across individual cells, we employed AUCell algorithm, which computes the Area Under the Curve (AUC) for the cumulative expression of input gene sets per cell, allowing robust pathway activity estimation at single-cell resolution. The resulting AUC scores were visualized in low-dimensional space using UMAP embeddings, and further analyzed across clusters and patient subgroups. Functional enrichment and pathway analyses were conducted to identify distinct transcriptional programs and signaling pathway activations among cancer cell subpopulations.\u003c/p\u003e\u003c/div\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eAOC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eArea over the curve\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eAUC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eArea under the curve\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eHSA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eHighest single agent\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eNOGR\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eNormalized organoid growth rate\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eNSCLC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eNon-small cell lung cancer\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ePCA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eprincipal component analysis\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ePDAC\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003ePancreatic ductal adenocarcinoma\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003ePDO\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003ePatient-derived organoid\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eRNA-seq\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eRNA sequencing\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003escRNA-seq\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eSingle cell RNA sequencing\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eUMAP\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eUniform manifold approximation and projection\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Ethics Committee of the University Hospital Antwerp (UZA) under protocols 17/30/339 and 14/47/480. Written informed consent was obtained from all participating patients in accordance with institutional and national ethical guidelines.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNA\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work is supported by: Research Foundation Flanders (FWO), Grant number; FWO-SB 1S27021N to M.L.C. and University Research Fund (BOF) of the University of Antwerp, Antwerp, Belgium, Grant number; FFB220225 to H.P. and S.S.. Part of the work was also supported by several donors, amongst which Willy Floren and Dedert Schilde vzw. The funders played no role in study design, data collection, analysis and interpretation of data, or the writing of this manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eM.L.C. conceived the study. M.L.C., F.R.F. and S.S. performed the experiments and data acquisition. G.R., V.H., J.H., P.V.S., and N.K. provided patient samples and clinical data. M.L.C. and S.S. analyzed the data and wrote the manuscript under the supervision of C.D.. H.V., M.P., E.S., F.L., H.P. and C.D. supervised the project and provided critical feedback throughout the study. All authors reviewed and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNA\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eSiegel RL, Kratzer TB, Giaquinto AN, Sung H, Jemal A. Cancer statistics, 2025. CA Cancer J Clin. 2025;75(1):10-45.\u003c/li\u003e\n \u003cli\u003eRamaswamy A, Srinivas S, Chaudhari V, Bhargava P, Bhandare M, Shrikhande SV, et al. Systemic therapy in pancreatic ductal adenocarcinomas (PDACs)-basis and current status. Ecancermedicalscience. 2022;16:1367.\u003c/li\u003e\n \u003cli\u003eZhang XP, Xu S, Gao YX, Zhao ZM, Zhao GD, Hu MG, et al. Early and late recurrence patterns of pancreatic ductal adenocarcinoma after pancreaticoduodenectomy: a multicenter study. Int J Surg. 2023;109(4):785-93.\u003c/li\u003e\n \u003cli\u003eIntegrated Genomic Characterization of Pancreatic Ductal Adenocarcinoma. Cancer Cell. 2017;32(2):185-203.e13.\u003c/li\u003e\n \u003cli\u003ede Langen AJ, Johnson ML, Mazieres J, Dingemans AC, Mountzios G, Pless M, et al. Sotorasib versus docetaxel for previously treated non-small-cell lung cancer with KRAS(G12C) mutation: a randomised, open-label, phase 3 trial. Lancet. 2023;401(10378):733-46.\u003c/li\u003e\n \u003cli\u003eJ\u0026auml;nne PA, Riely GJ, Gadgeel SM, Heist RS, Ou SI, Pacheco JM, et al. Adagrasib in Non-Small-Cell Lung Cancer Harboring a KRAS(G12C) Mutation. N Engl J Med. 2022;387(2):120-31.\u003c/li\u003e\n \u003cli\u003eWei D, Wang L, Zuo X, Maitra A, Bresalier RS. A Small Molecule with Big Impact: MRTX1133 Targets the KRASG12D Mutation in Pancreatic Cancer. Clin Cancer Res. 2024;30(4):655-62.\u003c/li\u003e\n \u003cli\u003eIsermann T, Sers C, Der CJ, Papke B. KRAS inhibitors: resistance drivers and combinatorial strategies. Trends in Cancer. 2025;11(2):91-116.\u003c/li\u003e\n \u003cli\u003eBrown WS, McDonald PC, Nemirovsky O, Awrey S, Chafe SC, Schaeffer DF, et al. Overcoming Adaptive Resistance to KRAS and MEK Inhibitors by Co-targeting mTORC1/2 Complexes in Pancreatic Cancer. Cell Reports Medicine. 2020;1(8):100131.\u003c/li\u003e\n \u003cli\u003eLiu X, Wang W, Liu X, Zhang Z, Yu L, Li R, et al. Multi-omics analysis of intra-tumoural and inter-tumoural heterogeneity in pancreatic ductal adenocarcinoma. Clinical and Translational Medicine. 2022;12(1):e670.\u003c/li\u003e\n \u003cli\u003eDriehuis E, van Hoeck A, Moore K, Kolders S, Francies HE, Gulersonmez MC, et al. Pancreatic cancer organoids recapitulate disease and allow personalized drug screening. Proc Natl Acad Sci U S A. 2019;116(52):26580-90.\u003c/li\u003e\n \u003cli\u003eHuang T, Cao H, Liu C, Sun X, Dai S, Liu L, et al. MAL2 reprograms lipid metabolism in intrahepatic cholangiocarcinoma via EGFR/SREBP-1 pathway based on single-cell RNA sequencing. Cell Death Dis. 2024;15(6):411.\u003c/li\u003e\n \u003cli\u003eWolf FA, Angerer P, Theis FJ. SCANPY: large-scale single-cell gene expression data analysis. Genome Biology. 2018;19(1):15.\u003c/li\u003e\n \u003cli\u003eWerba G, Weissinger D, Kawaler EA, Zhao E, Kalfakakou D, Dhara S, et al. Single-cell RNA sequencing reveals the effects of chemotherapy on human pancreatic adenocarcinoma and its tumor microenvironment. Nat Commun. 2023;14(1):797.\u003c/li\u003e\n \u003cli\u003eWang X, Allen S, Blake JF, Bowcut V, Briere DM, Calinisan A, et al. Identification of MRTX1133, a Noncovalent, Potent, and Selective KRASG12D Inhibitor. Journal of Medicinal Chemistry. 2022;65(4):3123-33.\u003c/li\u003e\n \u003cli\u003ePetrenko O, Kirillov V, D\u0026apos;Amico S, Reich NC. Intratumor heterogeneity in KRAS signaling shapes treatment resistance. iScience. 2025;28(2):111662.\u003c/li\u003e\n \u003cli\u003eMcDaid WJ, Wilson L, Adderley H, Martinez-Lopez A, Baker MJ, Searle J, et al. The PI3K-AKT-mTOR axis persists as a therapeutic dependency in KRASG12D-driven non-small cell lung cancer. Molecular Cancer. 2024;23(1):253.\u003c/li\u003e\n \u003cli\u003eIhle NT, Byers LA, Kim ES, Saintigny P, Lee JJ, Blumenschein GR, et al. Effect of KRAS oncogene substitutions on protein behavior: implications for signaling and clinical outcome. J Natl Cancer Inst. 2012;104(3):228-39.\u003c/li\u003e\n \u003cli\u003eHallin J, Bowcut V, Calinisan A, Briere DM, Hargis L, Engstrom LD, et al. Anti-tumor efficacy of a potent and selective non-covalent KRASG12D inhibitor. Nature Medicine. 2022;28(10):2171-82.\u003c/li\u003e\n \u003cli\u003eLee JE, Woo MG, Jung KH, Kang YW, Shin SM, Son MK, et al. Combination Therapy of the Active KRAS-Targeting Antibody inRas37 and a PI3K Inhibitor in Pancreatic Cancer. Biomol Ther (Seoul). 2022;30(3):274-83.\u003c/li\u003e\n \u003cli\u003eMisale S, Fatherree JP, Cortez E, Li C, Bilton S, Timonina D, et al. KRAS G12C NSCLC Models Are Sensitive to Direct Targeting of KRAS in Combination with PI3K Inhibition. Clin Cancer Res. 2019;25(2):796-807.\u003c/li\u003e\n \u003cli\u003eSuh KJ, Ryu MH, Zang DY, Bae WK, Lee HS, Oh HJ, et al. AZD8186 in Combination With Paclitaxel in Patients With Advanced Gastric Cancer: Results From a Phase Ib/II Study (KCSG ST18-20). Oncologist. 2023;28(9):e823-e34.\u003c/li\u003e\n \u003cli\u003eChoudhury AD, Higano CS, de Bono JS, Cook N, Rathkopf DE, Wisinski KB, et al. A Phase I Study Investigating AZD8186, a Potent and Selective Inhibitor of PI3K\u0026beta;/\u0026delta;, in Patients with Advanced Solid Tumors. Clin Cancer Res. 2022;28(11):2257-69.\u003c/li\u003e\n \u003cli\u003eMarqu\u0026eacute;s M, Tranchant R, Risa-Ebr\u0026iacute; B, Su\u0026aacute;rez-Sol\u0026iacute;s ML, Fern\u0026aacute;ndez LC, Carrillo-de-Santa-Pau E, et al. Combined MEK and PI3K/p110\u0026beta; Inhibition as a Novel Targeted Therapy for Malignant Mesothelioma Displaying Sarcomatoid Features. Cancer Res. 2020;80(4):843-56.\u003c/li\u003e\n \u003cli\u003eHwangbo H, Patterson SC, Dai A, Plana D, Palmer AC. Additivity predicts the efficacy of most approved combination therapies for advanced cancer. Nature Cancer. 2023;4(12):1693-704.\u003c/li\u003e\n \u003cli\u003eNing C, Liang M, Liu S, Wang G, Edwards H, Xia Y, et al. Targeting ERK enhances the cytotoxic effect of the novel PI3K and mTOR dual inhibitor VS-5584 in preclinical models of pancreatic cancer. Oncotarget. 2017;8(27):44295-311.\u003c/li\u003e\n \u003cli\u003eDolly SO, Wagner AJ, Bendell JC, Kindler HL, Krug LM, Seiwert TY, et al. Phase I Study of Apitolisib (GDC-0980), Dual Phosphatidylinositol-3-Kinase and Mammalian Target of Rapamycin Kinase Inhibitor, in Patients with Advanced Solid Tumors. Clin Cancer Res. 2016;22(12):2874-84.\u003c/li\u003e\n \u003cli\u003ePowles T, Lackner MR, Oudard S, Escudier B, Ralph C, Brown JE, et al. Randomized Open-Label Phase II Trial of Apitolisib (GDC-0980), a Novel Inhibitor of the PI3K/Mammalian Target of Rapamycin Pathway, Versus Everolimus in Patients With Metastatic Renal Cell Carcinoma. J Clin Oncol. 2016;34(14):1660-8.\u003c/li\u003e\n \u003cli\u003eMakker V, Recio FO, Ma L, Matulonis UA, Lauchle JO, Parmar H, et al. A multicenter, single-arm, open-label, phase 2 study of apitolisib (GDC-0980) for the treatment of recurrent or persistent endometrial carcinoma (MAGGIE study). Cancer. 2016;122(22):3519-28.\u003c/li\u003e\n \u003cli\u003eKim MJ, Lee SJ, Ryu JH, Kim SH, Kwon IC, Roberts TM. Combination of KRAS gene silencing and PI3K inhibition for ovarian cancer treatment. J Control Release. 2020;318:98-108.\u003c/li\u003e\n \u003cli\u003eYamamoto N, Fujiwara Y, Tamura K, Kondo S, Iwasa S, Tanabe Y, et al. Phase Ia/Ib study of the pan-class I PI3K inhibitor pictilisib (GDC-0941) administered as a single agent in Japanese patients with solid tumors and in combination in Japanese patients with non-squamous non-small cell lung cancer. Invest New Drugs. 2017;35(1):37-46.\u003c/li\u003e\n \u003cli\u003eSarker D, Ang JE, Baird R, Kristeleit R, Shah K, Moreno V, et al. First-in-human phase I study of pictilisib (GDC-0941), a potent pan-class I phosphatidylinositol-3-kinase (PI3K) inhibitor, in patients with advanced solid tumors. Clin Cancer Res. 2015;21(1):77-86.\u003c/li\u003e\n \u003cli\u003eShapiro GI, LoRusso P, Kwak E, Pandya S, Rudin CM, Kurkjian C, et al. Phase Ib study of the MEK inhibitor cobimetinib (GDC-0973) in combination with the PI3K inhibitor pictilisib (GDC-0941) in patients with advanced solid tumors. Invest New Drugs. 2020;38(2):419-32.\u003c/li\u003e\n \u003cli\u003eSch\u0026ouml;ffski P, Cresta S, Mayer IA, Wildiers H, Damian S, Gendreau S, et al. A phase Ib study of pictilisib (GDC-0941) in combination with paclitaxel, with and without bevacizumab or trastuzumab, and with letrozole in advanced breast cancer. Breast Cancer Res. 2018;20(1):109.\u003c/li\u003e\n \u003cli\u003eSoria JC, Adjei AA, Bahleda R, Besse B, Ferte C, Planchard D, et al. A phase IB dose-escalation study of the safety and pharmacokinetics of pictilisib in combination with either paclitaxel and carboplatin (with or without bevacizumab) or pemetrexed and cisplatin (with or without bevacizumab) in patients with advanced non-small cell lung cancer. Eur J Cancer. 2017;86:186-96.\u003c/li\u003e\n \u003cli\u003eKrop IE, Mayer IA, Ganju V, Dickler M, Johnston S, Morales S, et al. Pictilisib for oestrogen receptor-positive, aromatase inhibitor-resistant, advanced or metastatic breast cancer (FERGI): a randomised, double-blind, placebo-controlled, phase 2 trial. Lancet Oncol. 2016;17(6):811-21.\u003c/li\u003e\n \u003cli\u003eLeong S, Moss RA, Bowles DW, Ware JA, Zhou J, Spoerke JM, et al. A Phase I Dose-Escalation Study of the Safety and Pharmacokinetics of Pictilisib in Combination with Erlotinib in Patients with Advanced Solid Tumors. Oncologist. 2017;22(12):1491-9.\u003c/li\u003e\n \u003cli\u003eDilly J, Hoffman MT, Abbassi L, Li Z, Paradiso F, Parent BD, et al. Mechanisms of Resistance to Oncogenic KRAS Inhibition in Pancreatic Cancer. Cancer Discov. 2024;14(11):2135-61.\u003c/li\u003e\n \u003cli\u003eAndr\u0026eacute; F, Ciruelos EM, Juric D, Loibl S, Campone M, Mayer IA, et al. Alpelisib plus fulvestrant for PIK3CA-mutated, hormone receptor-positive, human epidermal growth factor receptor-2-negative advanced breast cancer: final overall survival results from SOLAR-1. Ann Oncol. 2021;32(2):208-17.\u003c/li\u003e\n \u003cli\u003eJhaveri KL, Im S-A, Saura C, Loibl S, Kalinsky K, Schmid P, et al. Overall Survival with Inavolisib in \u0026lt;i\u0026gt;PIK3CA\u0026lt;/i\u0026gt;-Mutated Advanced Breast Cancer. New England Journal of Medicine. 2025;393(2):151-61.\u003c/li\u003e\n \u003cli\u003eDienstmann R, Rodon J, Serra V, Tabernero J. Picking the Point of Inhibition: A Comparative Review of PI3K/AKT/mTOR Pathway Inhibitors. Molecular Cancer Therapeutics. 2014;13(5):1021-31.\u003c/li\u003e\n \u003cli\u003eLe Compte M, De La Hoz EC, Peeters S, Fortes FR, Hermans C, Domen A, et al. Single-organoid analysis reveals clinically relevant treatment-resistant and invasive subclones in pancreatic cancer. NPJ Precis Oncol. 2023;7(1):128.\u003c/li\u003e\n \u003cli\u003eDeben C, De La Hoz EC, Compte ML, Van Schil P, Hendriks JMH, Lauwers P, et al. OrBITS: label-free and time-lapse monitoring of patient derived organoids for advanced drug screening. Cell Oncol (Dordr). 2023;46(2):299-314.\u003c/li\u003e\n \u003cli\u003eLe Compte M, Cardenas De La Hoz E, Peeters S, Smits E, Lardon F, Roeyen G, et al. Multiparametric Tumor Organoid Drug Screening Using Widefield Live-Cell Imaging for Bulk and Single-Organoid Analysis: 1940-087X; 2022. e64434 p.\u003c/li\u003e\n \u003cli\u003eDeben C, Cardenas De La Hoz E, Rodrigues Fortes F, Le Compte M, Seghers S, Vanlanduit S, et al. Development and validation of the Normalized Organoid Growth Rate (NOGR) metric in brightfield imaging-based assays. Communications Biology. 2024;7(1):1612.\u003c/li\u003e\n \u003cli\u003eZheng S, Wang W, Aldahdooh J, Malyutina A, Shadbahr T, Tanoli Z, et al. SynergyFinder plus: toward better interpretation and annotation of drug combination screening datasets. Genomics, Proteomics and Bioinformatics. 2022;20(3):587-96.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Pancreatic ductal adenocarcinoma, KRAS G12D, MRTX1133, patient-derived organoids, single-cell RNA sequencing, combination therapies, drug synergy, PI3K–mTOR inhibition, targeted therapy resistance, functional precision oncology","lastPublishedDoi":"10.21203/rs.3.rs-7614104/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7614104/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eKRAS G12D is the most common oncogenic mutation in pancreatic ductal adenocarcinoma (PDAC), driving resistance and heterogeneity. Using robotic high-throughput drug screening with live-cell imaging, we evaluated normalized organoid growth rate (NOGR) in KRAS G12D-mutant PDAC organoids treated with the inhibitor MRTX1133. Monotherapy induced predominantly cytostatic, dose-dependent effects, reflecting heterogeneity observed in single-cell transcriptomics, where compensatory MAPK/ERK and PI3K\u0026ndash;mTOR activation emerged. Drug response profiles were variable, and synergy screening revealed patient- and compound-specific interactions. The most consistent cytotoxic synergy was achieved with PI3K\u0026ndash;mTOR inhibitors. Isoform-specific PI3Kα inhibitors (inavolisib, alpelisib) demonstrated robust synergy with MRTX1133 and favorable tumor selectivity, whereas dual PI3K/mTOR inhibitors (e.g., VS-5584) were more cytotoxic but lacked specificity. These results indicate that KRAS G12D inhibition alone is insufficient due to underlying transcriptional diversity, and highlight isoform-specific PI3Kα inhibitors as promising partners for combination therapy in PDAC.\u003c/p\u003e","manuscriptTitle":"Pathway-selective strategies reveal specific PI3K–mTOR inhibitors as key partners in KRAS G12D-targeted therapy","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-06 12:27:52","doi":"10.21203/rs.3.rs-7614104/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"29739df0-b40f-4f19-aaa8-5cd92bb13991","owner":[],"postedDate":"October 6th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":55740800,"name":"Biological sciences/Cancer"},{"id":55740801,"name":"Biological sciences/Drug discovery"},{"id":55740802,"name":"Health sciences/Oncology"}],"tags":[],"updatedAt":"2025-10-15T20:53:24+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-06 12:27:52","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7614104","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7614104","identity":"rs-7614104","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}