Quantitative Systems Pharmacology Model Research Articles

An integrated quantitative systems pharmacology virtual population approach for calibration with oncology efficacy endpoints.

In drug development, quantitative systems pharmacology (QSP) models are becoming an increasingly important mathematical tool for understanding response variability and for generating predictions to inform development decisions. Virtual populations are essential for sampling uncertainty and potential variability in QSP model predictions, but many clinical efficacy endpoints can be difficult to capture with QSP models that typically rely on mechanistic biomarkers. In oncology, challenges are particularly significant when connecting tumor size with time-to-event endpoints like progression-free survival while also accounting for censoring due to consent withdrawal, loss in follow-up, or safety criteria. Here, we expand on our prior work and propose an extended virtual population selection algorithm that can jointly match tumor burden dynamics and progression-free survival times in the presence of censoring. We illustrate the core components of our algorithm through simulation and calibration of a signaling pathway model that was fitted to clinical data for a small molecule targeted inhibitor. This methodology provides an approach that can be tailored to other virtual population simulations aiming to match survival endpoints for solid-tumor clinical datasets.

Understanding how CD19 expression levels impact the response to loncastuximab tesirine: a plain language summary.

In this article, we summarize results from a clinical study called LOTIS-2, in which researchers looked at patients with a type of blood cancer called diffuse large B-cell lymphoma, or DLBCL for short. Patients received the drug loncastuximab tesirine, or Lonca for short, which targets a marker on the surface of tumor cells called CD19.Patient information from the LOTIS-2 study, other studies of Lonca, and information from scientific publications was used to develop a quantitative systems pharmacology (QSP) model, which can predict how Lonca works in the body. The goal was to use the QSP model to see if CD19 levels can predict tumor size changes after Lonca treatment and if Lonca can still work to treat DLBCL when CD19 levels are very low. The prior LOTIS-1 and LOTIS-2 trials demonstrated an acceptable safety profile for Lonca, and therefore the current study did not evaluate safety data. Researchers used immunohistochemistry, a common technique to evaluate CD19 expression. They found that there was no association between patients who responded to Lonca treatment and levels of CD19 on their tumor cells. Some patients with low or even undetectable levels of CD19 on their tumor cells had observable decreases in tumor size after Lonca treatment. While Lonca uses the CD19 target to find and destroy cancer cells, Lonca does not require a large amount of CD19 to kill tumor cells. These results mean that Lonca may be an effective treatment for patients with DLBCL, even if CD19 expression in tumors is undetectable by immunohistochemistry.

Virtual patient analysis identifies strategies to improve the performance of predictive biomarkers for PD-1 blockade

Patients with metastatic triple-negative breast cancer (TNBC) show variable responses to PD-1 inhibition. Efficient patient selection by predictive biomarkers would be desirable but is hindered by the limited performance of existing biomarkers. Here, we leveraged in silico patient cohorts generated using a quantitative systems pharmacology model of metastatic TNBC, informed by transcriptomic and clinical data, to explore potential ways to improve patient selection. We evaluated and quantified the performance of 90 biomarker candidates, including various cellular and molecular species, at different cutoffs by a cutoff-based biomarker testing algorithm combined with machine learning-based feature selection. Combinations of pretreatment biomarkers improved the specificity compared to single biomarkers at the cost of reduced sensitivity. On the other hand, early on-treatment biomarkers, such as the relative change in tumor diameter from baseline measured at two weeks after treatment initiation, achieved remarkably higher sensitivity and specificity. Further, blood-based biomarkers had a comparable ability to tumor- or lymph node-based biomarkers in identifying a subset of responders, potentially suggesting a less invasive way for patient selection.

Landscape of regulatory quantitative systems pharmacology submissions to the U.S. Food and Drug Administration: An update report.

The number of quantitative systems pharmacology (QSP) submissions to the U.S. Food and Drug Administration has continued to increase over the past decade. This report summarizes the landscape of QSP submissions as of December 2023. QSP was used to inform drug development across various therapeutic areas and throughout the drug development process of small molecular drugs and biologics and has facilitated dose finding, dose ranging, and dose optimization studies. Though the majority of QSP submissions (>66%) focused on drug effectiveness, QSP was also utilized to simulate drug safety including liver toxicity, risk of cytokine release syndrome (CRS), bone density, and others. This report also includes individual contexts of use from a handful of new drug applications (NDAs) and biologics license applications where QSP modeling was used to demonstrate the utility of QSP modeling in regulatory drug development. According to the models submitted in QSP submissions, an anonymous case was utilized to illustrate how QSP informed development of a bispecific monoclonal antibody with respect to CRS risk. QSP submissions for informing pediatric drug development were summarized along with highlights of a case in inborn errors of metabolism. Furthermore, simulations of response variability with QSP were described. In summary, QSP continues to play a role in informing drug development.

Recommendations for a standardized publication protocol for a QSP model.

Development of a Quantitative Systems Pharmacology (QSP) model is a long process with many iterative steps. Lack of standard practices for publishing QSP models has resulted in limited model reproducibility within the field. Multiple studies have identified that model reproducibility is a large challenge, especially for QSP models. This work aimed to investigate the causes of QSP model reproducibility issues and suggest standard practices as a potential solution to ensure QSP models are reproducible. In addition, a protocol is suggested as a guidance towards better publication strategy across journals, hoping to enable QSP knowledge preservation.

Quantitative systems pharmacology model of α-synuclein pathology in Parkinson's disease-like mouse for investigation of passive immunotherapy mechanisms.

The main pathophysiological hallmark of Parkinson's disease (PD) is the accumulation of aggregated alpha-synuclein (αSyn). Microglial activation is an early event in PD and may play a key role in pathological αSyn aggregation and transmission, as well as in clearance of αSyn and immunotherapy efficacy. Our aim was to investigate how different proposed mechanisms of anti-αSyn immunotherapy may contribute to pathology reduction in various PD-like mouse models. Our mechanistic model of PD pathology in mouse includes αSyn production, aggregation, degradation and distribution in neurons, secretion into interstitial fluid, internalization, and subsequent clearance by neurons and microglia. It describes the influence of neuroinflammation on PD pathogenesis and dopaminergic neurodegeneration. Multiple data from mouse PD models were used for calibration and validation. Simulations of anti-αSyn passive immunotherapy adequately reproduce preclinical data and suggest that (1) immunotherapy is efficient in the reduction of aggregated αSyn in various models of PD-like pathology; (2) prevention of aSyn spread only does not reduce the pathology; (3) a decrease in microglial inflammatory activation and aSyn aggregation may be alternative therapy approaches in PD-like pathology.

Identifying biomarkers for treatment of uveal melanoma by T cell engager using a QSP model.

Uveal melanoma (UM), the primary intraocular tumor in adults, arises from eye melanocytes and poses a significant threat to vision and health. Despite its rarity, UM is concerning due to its high potential for liver metastasis, resulting in a median survival of about a year after detection. Unlike cutaneous melanoma, UM responds poorly to immune checkpoint inhibition (ICI) due to its low tumor mutational burden and PD-1/PD-L1 expression. Tebentafusp, a bispecific T cell engager (TCE) approved for metastatic UM, showed potential in clinical trials, but the objective response rate remains modest. To enhance TCE efficacy, we explored quantitative systems pharmacology (QSP) modeling in this study. By integrating a TCE module into an existing QSP model and using clinical data on UM and tebentafusp, we aimed to identify and rank potential predictive biomarkers for patient selection. We selected 30 important predictive biomarkers, including model parameters and cell concentrations in tumor and blood compartments. We investigated biomarkers using different methods, including comparison of median levels in responders and non-responders, and a cutoff-based biomarker testing algorithm. CD8+ T cell density in the tumor and blood, CD8+ T cell to regulatory T cell ratio in the tumor, and naïve CD4+ density in the blood are examples of key biomarkers identified. Quantification of predictive power suggested a limited predictive power for single pre-treatment biomarkers, which was improved by early on-treatment biomarkers and combination of predictive biomarkers. Ultimately, this QSP model could facilitate biomarker-guided patient selection, improving clinical trial efficiency and UM treatment outcomes.

Quantitative Systems Pharmacology Models: Potential Tools for Advancing Drug Development for Rare Diseases.

Rare diseases, affecting millions globally, present significant drug development challenges. This is due to the limited patient populations and the unique pathophysiology of these diseases, which can make traditional clinical trial designs unfeasible. Quantitative Systems Pharmacology (QSP) models offer a promising approach to expedite drug development, particularly in rare diseases. QSP models provide a mechanistic representation of the disease and drug response in virtual patients that can complement routinely applied empirical modeling and simulation approaches. QSP models can generate digital twins of actual patients and mechanistically simulate the disease progression of rare diseases, accounting for phenotypic heterogeneity. QSP models can also support drug development in various drug modalities, such as gene therapy. Impactful QSP models case studies are presented here to illustrate their value in supporting various aspects of drug development in rare indications. As these QSP model applications continue to mature, there is a growing possibility that they could be more widely integrated into routine drug development steps. This integration could provide a robust framework for addressing some of the inherent challenges in rare disease drug development.

ISB 2001 trispecific T cell engager shows strong tumor cytotoxicity and overcomes immune escape mechanisms of multiple myeloma cells

Despite recent advances in immunotherapies targeting single tumor-associated antigens, patients with multiple myeloma eventually relapse. ISB 2001 is a CD3+ T cell engager (TCE) co-targeting BCMA and CD38 designed to improve cytotoxicity against multiple myeloma. Targeting of two tumor-associated antigens by a single TCE resulted in superior cytotoxic potency across a variable range of BCMA and CD38 tumor expression profiles mimicking natural tumor heterogeneity, improved resistance to competing soluble factors and exhibited superior cytotoxic potency on patient-derived samples and in mouse models. Despite the broad expression of CD38 across human tissues, ISB 2001 demonstrated a reduced T cell activation profile in the absence of tumor cells when compared to TCEs targeting CD38 only. To determine an optimal first-in-human dose for the ongoing clinical trial (NCT05862012), we developed an innovative quantitative systems pharmacology model leveraging preclinical data, using a minimum pharmacologically active dose approach, therefore reducing patient exposure to subefficacious doses of therapies.

ABCL-366 Quantitative Systems Pharmacology Model Predicts Combination Activity of CD19-Targeted Loncastuximab Tesirine Co-Dosed With a CD20/CD3 T-Cell Bispecific (Epcoritamab) in Diffuse Large B-Cell Lymphoma

ABCL-366 Quantitative Systems Pharmacology Model Predicts Combination Activity of CD19-Targeted Loncastuximab Tesirine Co-Dosed With a CD20/CD3 T-Cell Bispecific (Epcoritamab) in Diffuse Large B-Cell Lymphoma

Quantitative Systems Pharmacology Model Predicts Combination Activity of CD19-Targeted Loncastuximab Tesirine Co-Dosed With a CD20/CD3 T-Cell Bispecific (Epcoritamab) in Diffuse Large B-Cell Lymphoma

Quantitative Systems Pharmacology Model Predicts Combination Activity of CD19-Targeted Loncastuximab Tesirine Co-Dosed With a CD20/CD3 T-Cell Bispecific (Epcoritamab) in Diffuse Large B-Cell Lymphoma

Real-World Application of a Quantitative Systems Pharmacology (QSP) Model to Predict Potassium Concentrations from Electronic Health Records: A Pilot Case towards Prescribing Monitoring of Spironolactone.

Quantitative systems pharmacology (QSP) models are rarely applied prospectively for decision-making in clinical practice. We therefore aimed to operationalize a QSP model for potas-sium homeostasis to predict potassium trajectories based on spironolactone administrations. For this purpose, we proposed a general workflow that was applied to electronic health records (EHR) from patients treated in a German tertiary care hospital. The workflow steps included model exploration, local and global sensitivity analyses (SA), identifiability analysis (IA) of model parameters, and specification of their inter-individual variability (IIV). Patient covariates, selected parameters, and IIV then defined prior information for the Bayesian a posteriori prediction of individual potassium trajectories of the following day. Following these steps, the successfully operationalized QSP model was interactively explored via a Shiny app. SA and IA yielded five influential and estimable parameters (extracellular fluid volume, hyperaldosteronism, mineral corticoid receptor abundance, potassium intake, sodium intake) for Bayesian prediction. The operationalized model was validated in nine pilot patients and showed satisfactory performance based on the (absolute) average fold error. This provides proof-of-principle for a Prescribing Monitoring of potassium concentrations in a hospital system, which could suggest preemptive clinical measures and therefore potentially avoid dangerous hyperkalemia or hypokalemia.

Challenges with sirolimus experimental data to inform QSP model of post-transplantation cyclophosphamide regimens.

Dose optimization of sirolimus may further improve outcomes in allogeneic hematopoietic cell transplant (HCT) patients receiving post-transplantation cyclophosphamide (PTCy) to prevent graft-versus-host disease (GVHD). Sirolimus exposure-response association studies in HCT patients (i.e., the association of trough concentration with clinical outcomes) have been conflicting. Sirolimus has important effects on T-cells, including conventional (Tcons) and regulatory T-cells (Tregs), both of which have been implicated in the mechanisms by which PTCy prevents GVHD, but there is an absence of validated biomarkers of sirolimus effects on these cell subsets. Considering the paucity of existing biomarkers and the complexities of the immune system, we conducted a literature review to inform a quantitative systems pharmacology (QSP) model of GVHD. The published literature presented multiple challenges. The sirolimus pharmacokinetic models insufficiently describe sirolimus distribution to relevant physiological compartments. Despite multiple publications describing sirolimus effects on Tcons and Tregs in preclinical and human exvivo models, consistent parameters relating sirolimus concentrations to circulating Tcons and Tregs could not be found. Each aspect presents a challenge in building a QSP model of sirolimus and its temporal effects on T-cell subsets and GVHD prevention. To optimize GVHD prevention regimens, phase I studies and systematic studies of immunosuppressant concentration-effect association are needed for QSP modeling.

Life scientists improve QSP model quality and impact.

Life scientists improve QSP model quality and impact.

Analysis of Cerebrospinal Fluid, Plasma β-Amyloid Biomarkers, and Cognition from a 2-Year Phase 2 Trial Evaluating Oral ALZ-801/Valiltramiprosate in APOE4 Carriers with Early Alzheimer's Disease Using Quantitative Systems Pharmacology Model.

ALZ-801/valiltramiprosate is an oral, small-molecule inhibitor of beta-amyloid (Aβ) aggregation and oligomer formation in late-stage development as a disease-modifying therapy for early Alzheimer's disease (AD). The present investigation provides a quantitative systems pharmacology (QSP) analysis of amyloid fluid biomarkers and cognitive results from a 2-year ALZ-801 Phase 2 trial in APOE4 carriers with early AD. The single-arm, open-label phase 2 study evaluated effects of ALZ-801 265 mg two times daily (BID) on cerebrospinal fluid (CSF) and plasma amyloid fluid biomarkers over 104 weeks in APOE4 carriers with early AD [Mini-Mental State Examination (MMSE) ≥ 22]. Subjects with positive CSF biomarkers for amyloid (Aβ42/Aβ40) and tau pathology (p-tau181) were enrolled, with serial CSF and plasma levels of Aβ42 and Aβ40 measured over 104 weeks. Longitudinal changes of CSF Aβ42, plasma Aβ42/Aβ40 ratio, and cognitive Rey Auditory Verbal Learning Test (RAVLT) were compared with the established natural disease trajectories in AD using a QSP approach. The natural disease trajectory data for amyloid biomarkers and RAVLT were extracted from a QSP model and an Alzheimer's disease neuroimaging initiative population model, respectively. Analyses were stratified by disease severity and sex. A total of 84 subjects were enrolled. Excluding one subject who withdrew at the early stage of the trial, data from 83 subjects were used for this analysis. The ALZ-801 treatment arrested the progressive decline in CSF Aβ42 level and plasma Aβ42/Aβ40 ratio, and stabilized RAVLT over 104 weeks. Both sexes showed comparable responses to ALZ-801, whereas mild cognitive impairment (MCI) subjects (MMSE ≥ 27) exhibited a larger biomarker response compared with more advanced mild AD subjects (MMSE 22-26). In this genetically defined and biomarker-enriched early AD population, the QSP analysis demonstrated a positive therapeutic effect of oral ALZ-801 265 mg BID by arresting the natural decline of monomeric CSF and plasma amyloid biomarkers, consistent with the target engagement to prevent their aggregation into soluble neurotoxic oligomers and subsequently into insoluble fibrils and plaques over 104 weeks. Accompanying the amyloid biomarker changes, ALZ-801 also stabilized the natural trajectory decline of the RAVLT memory test, suggesting that the clinical benefits are consistent with its mechanism of action. This sequential effect arresting the disease progression on biomarkers and cognitive decline was more pronounced in the earlier symptomatic stages of AD. The QSP analysis provides fluid biomarker and clinical evidence for ALZ-801 as a first-in-class, oral small-molecule anti-Aβ oligomer agent with disease modification potential in AD. https://clinicaltrials.gov/study/NCT04693520.

Virtual clinical trials via a QSP immuno-oncology model to simulate the response to a conditionally activated PD-L1 targeting antibody in NSCLC

AbstractRecently, immunotherapies for antitumoral response have adopted conditionally activated molecules with the objective of reducing systemic toxicity. Amongst these are conditionally activated antibodies, such as PROBODY® activatable therapeutics (Pb-Tx), engineered to be proteolytically activated by proteases found locally in the tumor microenvironment (TME). These PROBODY® therapeutics molecules have shown potential as PD-L1 checkpoint inhibitors in several cancer types, including both effectiveness and locality of action of the molecule as shown by several clinical trials and imaging studies. Here, we perform an exploratory study using our recently published quantitative systems pharmacology model, previously validated for triple-negative breast cancer (TNBC), to computationally predict the effectiveness and targeting specificity of a PROBODY® therapeutics drug compared to the non-modified antibody. We begin with the analysis of anti-PD-L1 immunotherapy in non-small cell lung cancer (NSCLC). As a first contribution, we have improved previous virtual patient selection methods using the omics data provided by the iAtlas database portal compared to methods previously published in literature. Furthermore, our results suggest that masking an antibody maintains its efficacy while improving the localization of active therapeutic in the TME. Additionally, we generalize the model by evaluating the dependence of the response to the tumor mutational burden, independently of cancer type, as well as to other key biomarkers, such as CD8/Treg Tcell and M1/M2 macrophage ratio. While our results are obtained from simulations on NSCLC, our findings are generalizable to other cancer types and suggest that an effective and highly selective conditionally activated PROBODY® therapeutics molecule is a feasible option.

Quantitative systems pharmacology (QSP) model to predict combination activity of CD19-targeted antibody drug conjugate (loncastuximab tesirine) co-dosed with a CD20/CD3 T-cell bispecific (epcoritamab) in patients with diffuse large B-cell lymphoma (DLBCL).

e19059 Background: Loncastuximab tesirine-lpy (Lonca) is an antibody-drug conjugate comprising an anti-CD19 antibody conjugated to a pyrrolobenzodiazepine dimer cytotoxin. Epcoritamab-bysp (Epco) is a CD20×CD3 T cell-engaging bispecific antibody (bsAb) that redirects T-cells to eliminate malignant B-cells. Epco and Lonca target different B-cell surface antigens and combining both is expected to have additive or synergistic efficacy. Previously, a novel QSP model for Lonca was developed (Utsey K et al. Clin Pharmacol Drug Dev. 2023;12:123-125) and validated with clinical data from patients with R/R DLBCL (Caimi P et al. Blood. 2022;140:9548-9550). A combination QSP model for Lonca with Epco in R/R DLBCL was described to predict potential anti-tumor synergies in proposed clinical trials. Methods: The model reduced previous Lonca model complexity while maintaining core functionality and was combined with a published QSP model for T-cell bsAbs (Hosseini I et al. Npj Syst Bio Appl. 2020;6(1):1-11) to predict tumor dynamics after Lonca and/or Epco treatment. Assumptions included: 1) Lonca or Epco induce healthy and malignant B-cell killing; 2) tumor comprises malignant B- & T-cells; 3) tumor volume is based on malignant B-cell count; 4) T-cells can enter or leave tumor as with other tissues; 5) CD19-/low CD20+ B-cells account for tumor heterogeneity from cells insensitive to Lonca treatment. Results: Tumor growth inhibition (TGI) was predicted to plateau by Cycle (C) 2 with Epco monotherapy. By end of C3, however, predicted median normalized tumor volume reduction with co-dosing Lonca and Epco was approximately 1 log order greater than Epco alone. Maximum activity of Lonca and Epco therapy was not observed until C4 or later as depth of tumor regression increased with additional treatment cycles. Since patients with DLBCL may present with low T-cell counts (Hutchings et al. Lancet. 2021) the effect of T-cell count at baseline was evaluated. A 50% decrease at baseline from 2000-1000 T-cells/µL significantly reduced antitumor effect with Epco alone but made no difference with Lonca in combination. Conclusions: These results resemble predictions for Lonca in combination with other bsAbs (Lonca plus mosunetuzumab or glofitamab; abstract submitted to AACR 2024). By end of C3, Lonca plus Epco was predicted to promote substantially more TGI than Epco or Lonca alone. While increased doses of Lonca from 90-150 µg/kg in combination had limited additional therapeutic benefit, combination treatment beyond C3 was predicted to increase depth of tumor regression. Response from Lonca and Epco co-dosing was predicted to be less affected by suppressed T-cell counts at baseline compared to Epco alone; a feature that may enhance responsiveness when T-cell level is moderate.

Quantitative systems pharmacology modeling of macrophage-targeted therapy combined with PD-L1 inhibition in advanced NSCLC.

Immune checkpoint inhibitors remained the standard-of-care treatment for advanced non-small cell lung cancer (NSCLC) for the past decade. In unselected patients, anti-PD-(L)1 monotherapy achieved an overall response rate of about 20%. In this analysis, we developed a pharmacokinetic and pharmacodynamic module for our previously calibrated quantitative systems pharmacology model (QSP) to simulate the effectiveness of macrophage-targeted therapies in combination with PD-L1 inhibition in advanced NSCLC. By conducting in silico clinical trials, the model confirmed that anti-CD47 treatment is not an optimal option of second- and later-line treatment for advanced NSCLC resistant to PD-(L)1 blockade. Furthermore, the model predicted that inhibition of macrophage recruitment, such as using CCR2 inhibitors, can potentially improve tumor size reduction when combined with anti-PD-(L)1 therapy, especially in patients who are likely to respond to anti-PD-(L)1 monotherapy and those with a high level of tumor-associated macrophages. Here, we demonstrate the application of the QSP platform on predicting the effectiveness of novel drug combinations involving immune checkpoint inhibitors based on preclinical or early-stage clinical trial data.

Optimizing Clinical Translation of Bispecific T-cell Engagers through Context Unification with a Quantitative Systems Pharmacology Model.

Bispecific T-cell engagers (bsTCEs) have emerged as a promising class of cancer immunotherapy. BsTCEs enable physical connections between T cells and tumor cells to enhance T-cell activity against cancer. Despite several marketing approvals, the development of bsTCEs remains challenging, especially at early clinical translational stages. The intricate design of bsTCEs makes their pharmacologic effects and safety profiles highly dependent on patient's immunological and tumor conditions. Such context-dependent pharmacology introduces considerable uncertainty into translational efforts. In this study, we developed a Quantitative Systems Pharmacology (QSP) model, through context unification, that can facilitate the translation of bsTCEs preclinical data into clinical activity. Through characterizing the formation dynamics of immunological synapse (IS) induced by bsTCEs, this model unifies a broad range of contexts related to target affinity, tumor characteristics, and immunological conditions. After rigorous calibration using both experimental and clinical data, the model enables consistent translation of drug potency observed under diverse experimental conditions into predictable exposure-response relationships in patients. Moreover, the model can help identify optimal target-binding affinities and minimum efficacious concentrations across different clinical contexts. This QSP approach holds significant promise for the future development of bsTCEs.

Quantitative systems pharmacology modeling of tumor heterogeneity in response to BH3-mimetics using virtual tumors calibrated with cell viability assays.

Both primary and acquired resistance mechanisms that involve intra-tumoral cell heterogeneity limit the use of BH3-mimetics to trigger tumor cell apoptosis. This article proposes a new quantitative systems pharmacology (QSP)-based methodology in which cell viability assays are used to calibrate virtual tumors (VTs) made of virtual cells whose fate is determined by simulations from an apoptosis QSP model. VTs representing SU-DHL-4 and KARPAS-422 cell lines were calibrated using invitro data involving venetoclax (anti-BCL2), A-1155463 (anti-BCLXL), and/or A-1210477 (anti-MCL1). The calibrated VTs provide insights into the combination of several BH3-mimetics, such as the distinction between cells eliminated by at least one of the drugs (monotherapies) from the cells eliminated by a pharmacological combination only. Calibrated VTs can also be used as initial conditions in an agent-based model (ABM) framework, and a minimal ABM was developed to bridge invitro SU-DHL-4 cell viability results to tumor growth inhibition experiments in mice.