No abstract

There is growing interest in the use of human tumor primary cultures for the prediction of response to drugs and drug combinations in cancer medicine. While the field has witnessed numerous attempts to apply human tissue for the prediction of response, fundamental errors in methodologic approaches severely limited the predictive validity and clinical applicability of these important technologies. The first error was a focus upon drug induced cell growth inhibition. This lead to the failure of the clonogenic (human tumor stem cell) and H3*Thymidine incorporation platforms in the 1970’s and 1980’s. As the concept of apoptosis arose following landmark observations in 1972, cell death endpoints became predominant including mitochondrial function (MTT, XTT), ATP content (luciferase) and delayed loss of membrane integrity (DISC, EVA/PCD) among others. These improvements led to more biologically relevant measures. However, a second hurdle arose as investigators chose to amplify subculture and propagate tumors into what are now referred to as human tumor organoids. While these 3-D cultures are more reflective of tumor cell-cell interactions found in cancers, they cannot recreate the cell-vasculature, cell-stroma, inflammatory environment nor cancer associated fibroblast interactions with human cancer cells required for accurate response prediction. To address these shortcomings we developed the ex vivo analysis of programmed cell death (EVA/PCD) platform that applies native state tumor explants directly from surgical specimens to select drugs and combinations. To date we have reported 11,680 individual analyses. Positive prospective results have been reported in breast cancer (TTP p < 0.03); ovarian cancer (TTP p = 0.05), lung cancer (2-fold improvement in ORR; TTP p =0.035) and recently ovarian and uterine cancer (DFS p = 0.038) with two meta-analyses confirming the predictive validity in hematologic and solid tumors. With over 20 years of CLIA-approved laboratory experience in the study of human tumor primary culture explants, the EVA/PCD platform provides improved response and time to progression and offers the opportunity to streamline drug development, curtail costs and avoid futile care in patients with advanced and refractory malignancies. Citation Format: Adam J. Nagourney, Steven S. Evans, Paulo D'Amora, Nise Yamaguchi, Paula Bernard, Federico Francisco, Joshua B. Gipoor, Robert Alan Nagourney. Human tumor primary culture analyses for the prediction of response to chemotherapy, targeted agents, drug combinations and metabolic inhibitors: The role of ex vivo analysis of programmed cell death in cancer therapy. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4910.

Ovarian cancer is the leading cause of gynecologic cancer death with over 21,000 new diagnosis and 13,000 deaths each year in the US. Treatment with platinum-based chemotherapy induces remission in up to 80% of patients yet the majority relapse and die from their disease within 5 years. We recently reported the analysis of patients with advanced ovarian cancers using plasma metabolomics to predict response and survival (D’Amora et al Gynecol Oncol, October 2021). As part of that study human tumor explants were submitted for the measurement of platinum resistance by ex vivo analysis of programmed cell death (EVA/PCD࣪). Media obtained from the non-drug exposed control cultures were examined by mass spectrometry for the quantification of 186 individual metabolites using the Biocrates’ p180 platform. Results obtained with the tissue culture media identify biochemical signatures that correlate with the achievement of complete pathologic remission. Among the most discrimination findings were the media concentrations of phosphatidyl choline (PC) aa c42:0 and the ratio of (PC) aa C40:2 divided by (PC) ae C42:1 suggesting either hyperactivity of lipid elongases or defective beta oxidation of very long chain fatty acids in drug refractory patients. When we compared the results of 8 patients who achieved pCR with 3 patients with residual disease the aa C42:0 concentration and the ratio aa C42:0/ae C42:1 provided an AUC= 0.958 (p. < 0.01) The findings suggest that peroxisomal dysfunction leads to the accumulation of very long chain fatty acid lipids. These species are known to induce endoplasmic reticulum and oxidative stress and are associated with the neurodegenerative diseases like Zellweger syndrome. We have reported similar findings in drug refractory breast cancer patients (da Silva et al, Oncotarget, August Aug 2018) suggesting common lipid metabolic changes in drug resistant adenocarcinomas. The results of these analyses could provide prognostic and predictive information for patients diagnosed with advanced epithelial ovarian cancers. Further studies correlating these findings with plasma metabolic signatures are underway as will be reported. Citation Format: Adam J. Nagourney, Paulo D'Amora, Ismael Dale Cotrim Guerreiro Silva, Steven S. Evans, Robert A. Nagourney. Complete remission in advanced epithelial ovarian cancer is predicted by metabolic signatures measured in the media of patient’s primary culture 3D tumor explants: A metabolomic analysis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1965.

Background: TP53 is the most commonly mutated gene in human cancer with loss of function mutations largely concentrated in “hotspots” affecting DNA binding. APR-246 and COTI-2 are small molecules under investigation in P53 mutated cancers. APR binds to P53 cysteine residues, altering conformation, while COTI-2 showed activity in P53 mutant tumors by a computational platform. We compared APR-246 and COTI-2 activity in human tumor explants from 247 surgical specimens. Methods: Ex vivo analyses of programmed cell death measured drug-induced cell death by delayed-loss-of-membrane integrity and ATP content. The LC50s were compared by Z-Score. Synergy was conducted by the method of Chou and Talalay, and correlations were performed by Pearson moment. Results: APR-246 and COTI-2 activity favored hematologic neoplasms, but solid tumor activity varied by diagnosis. COTI-2 and APR-246 activity did not correlate (R = 0.1028) (NS). COTI-2 activity correlated with nitrogen mustard, cisplatin and gemcitabine, doxorubicin and selumetinib, with a trend for APR-246 with doxorubicin. For ovarian cancer, COTI-2 showed synergy with cisplatin at 25%. Conclusions: COTI-2 and APR-246 activity differ by diagnosis. A lack of correlation supports distinct modes of action. Cisplatin synergy is consistent with P53’s role in DNA damage. Different mechanisms of action may underlie disease specificity and offer better disease targeting.

Phosphoinositide 3-kinases (PI3Ks) are hetero-dimeric lipid kinases that are frequently dysregulated in cancer. Class I PI3K activity is upregulated in many breast cancers via mutations in PIK3CA, PIK3R1, PTEN and AKT1 making PI3K inhibition attractive as a therapeutic target. BYL719, an oral α-specific PI3K inhibitor that selectively inhibits p110α was granted FDA approval for estrogen receptor (ER) + breast cancer following the successful completion of the SOLAR-1 trial (Andre, F. et al NEJM, 2019). To explore the clinical utility of BYL 719 in human cancer, we applied the Ex Vivo Analysis of Programmed Cell Death (EVA/PCD) in surgical specimens obtained from 388 individual patients. Methods: EVA/PCD technique uses enzymatic and mechanical disaggregation followed by density centrifugation to isolate human tumor 3-D organoids of desired size. Morphologic and metabolic measures of cell death provide lethal concentration 50% (LC50), interpolated from 5-point dose response curves and then compared by modified Z-score to provide disease-specific activity. BYL 719 correlations with other classes of drugs are performed by Pearson Moment with significance by two tailed T-test. Results: By rank order the highest BYL719 activity was found in Uterine>Breast>Colon>Gastric>Ovary> Pancreas. The lowest BYL719 activity was found for Melanoma<Heme<GBM<Neuroendocrine<Sarcoma<NSCLC. Drug correlations were significant (Bonferroni-correction) for Idelalisib, Pictilisib, BEZ 235 and Everolimus as well as Gefitinib, Lapatinib, Selumetinib and Palbociclib. Among cytotoxic agents tested BYL719 activity only correlated with CDDP and 5FU. Conclusions: Whilst BYL 719 has been FDA approved for ER (+) breast cancer, our results suggest a strong rationale for its use in uterine cancer. As uterine cancer is associated with mutations in PTEN and PI3K, the BYL719 activity identified by EVA/PCD is highly consistent with the known molecular profiles. In an analysis of 87 breast cancer specimens, by subgroup (ER, PR, HER2, BRCA, etc.) only previously untreated patients revealed significantly higher BYL 719 activity. Based upon mechanism of action, correlations with PIK3 and PTEN inhibitors are expected, but those with EGFR, HER2 and CK 4/6 inhibitors may offer insights into potential synergies currently being evaluated by EVA/PCD. Discussion: BYL719's approval marks the introduction of a new drug class into clinical therapy. With a growing number of small molecule signal transduction inhibitors, the capacity to target PIK3 offers the opportunity to forge new drug combinations. The EVA/PCD platform's capacity to examine disease-specific activity and drug synergy offers the opportunity to streamline drug development, accelerate the introduction of new classes of agents and explore novel drug combinations. Citation Format: Robert Alan Nagourney, Adam J. Nagourney, Peter Tran, Paula Bernard, Federico Francisco, Steven S. Evans. First-in-class approved phosphoinositide-3-kinase inhibitor Alpelesib (BYL 719): Examination of disease-specific activity in human tumor 3 dimensional primary culture organoids [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 675.