Oral: Lewis Sheiner Student Session

C-03 Dynamic Energy Budget (DEB) based models of tumor-in-host growth inhibition and cachexia onset

E. M. Tosca (1), M. Rocchetti (2), E. Pesenti (3), P. Magni (1)

(1) Dipartimento di Ingegneria Industriale e dell’Informazione, Università degli Studi di Pavia, I-27100 Pavia, Italy; (2) Consultant, Milano, Italy; (3) Accelera srl, Nerviano (MI), Italy.

Objectives: The great contribution of PK-PD tumor growth inhibition (TGI) models in the anticancer drug development process is already well-established. However, models currently available are always focused only on the drug efficacy assessment and, completely neglecting the host organism, overlook the drug/tumor-related toxic effects [1, 2][PM1] . Actually, host conditions significantly influence tumor growth that, in turn, has a relevant impact on the host. Severe body weight (BW) loss (cachexia) and reduced food intake (anorexia) are among the main causes of cancer death and, also, relevant endpoints in the preclinical studies. Finding the best compromise between efficacy and toxicity is the goal of any anticancer therapy. In absence of appropriate models that consider both the tumor and host body interactions (tumor-in-host models) and the anticancer drug effects, this efficacy/toxicity evaluation is based on heavy and time-consuming trial-and-error procedures. Here, a new modeling approach able to describe tumor-in-host growth dynamics and cachexia onset during an anticancer treatment is proposed to better exploit data routinely generated in the preclinical phase of an oncological drug development process.

Methods: Tumor-in-host DEB-based model: Following the van Leeuwen work [3], the Dynamic Energy Budget (DEB) theory [4] is adopted as general framework to describe the host organism. The dynamics of host body, composed by the structural biomass and the energy reserve, follow from an energy balance. Tumor is conceived as an additional component able to subtract a fraction k_u of the host energy for its maintenance and growth. As tumor exploits host resources, in certain conditions, the organism can even degrade its structural biomass to survive and, at the same time, to satisfy the tumor energy demand (tumor-related cachexia). This condition can be further worsened by the negative impact of tumor progression on host energy intake (tumor-related anorexia).

Tumor-in-host DEB-TGI models: The tumor-in-host DEB-based model is extended and adapted to describe the effects of different anticancer treatments. 1) Cytotoxic agents:  Drug exerts a direct killing effect on tumor cells, modeled as in the Simeoni model [5], and an inhibitory effect on the host assimilation. The latter accounts for the temporary decreased energy intake (drug-related anorexia) due to drug side effects and followed by host BW loss (drug-related cachexia). 2) Anti-angiogenic agents: An inhibitory effect, linked to the drug concentration, is added on k_u fraction to account for the modification of the energy partition between tumor and host that follows the reduction of tumor vascularization due to the anti-angiogenic therapy. 3) Combination of anti-angiogenics with chemotherapy: A joint model, incorporating both the anti-angiogenic and cytotoxic DEB-TGI models, is used to predict tumor and host response to a combination therapy under a ‘no-interaction’ assumption [6]. The nature of combination (additivity/synergisms/ antagonisms) can be evaluated comparing model predicted and observed tumor weight (TW) and host BW.

Experimental data refer to TW and net host BW of 16 xenograft mice studies involving 6 tumor cell lines and 14 anticancer agents administered at several doses and schedules [7,8,9]. Furthermore, a study assessing etoposide (ETO) effects on both tumor-free and tumor-bearing Wistar rats is considered [10]. Average and individual data are analysed in Monolix 2016R1, by a naïve average and a non-linear-mixed-effect approach, respectively.

Results: Tumor-in-host DEB-based model: The model, identified on control animals, successfully describes TW and host BW, predicting a S-shape tumor growth profile that directly follows from physiological hypothesis on tumor-host energetic interactions. Cytotoxic agents: For 8 xenograft studies involving 3 tumor lines, the model is able to simultaneously describe and predict TW and host BW growth in control and treated mice with both novel anticancer compounds and well-known drugs (paclitaxel, 5-FU, cisplatin, vincristine, vinblastine and gemcitabine) [7]. A slightly revised model formulation, combined with the use of intratumoral concentration as driver of tumor kinetics, successfully describes the ETO effects on Wistar rats accounting also for its schedule-dependence [10]. This well-design experiment, including treated and untreated tumor-free animals, allows to fully exploit model capabilities in describing and discerning all the in vivo growth dynamics. Anti-angiogenic agents: The tumor-in-host DEB-based TGI model, adapted for cytostatic therapy, is successfully applied to 7 xenograft mice experiments assessing the Bevacizumab (BVZ) effect on 3 tumor cell lines [8]. In this case, in addition to the drug potency estimates, quantitative measurements of tumor-related cachexia are provided. Finally, a hypoxia-triggered resistance model allows to describe the decreased BVZ efficacy observed after prolonged treatments [9]. Combination of anti-angiogenics with chemotherapy: A combination study on xenograft mice treated with BVZ, 2 doses of NMS-937H or a combination of both is successfully analysed [8]. Model parameters estimated on the single-agent arms are used to predict the expected tumor and host response in the combination groups. Comparing the predicted curves with the observed data, no significant departures from additivity are found for both efficacy (TGI) and safety (cachexia) profiles. However, an increment in the TGI due to BVZ and NMS-937H coadministration highlights the advantage of the combination strategy.

Conclusions: A simultaneous modeling of tumor and host organism interactions during anticancer treatments is proposed on the basis of the DEB theory. This approach, suitably adapted to several preclinical experimental contexts, is able to integrate all the different aspects characterizing the in vivo TGI studies: drug cytotoxic or cytostatic activity on tumor, onset of drug/tumor-related cachexia and anorexia and influence of host condition on tumor growth. This allows for the first time to investigate separately BW loss due to tumor progression and to treatment, providing in addition better estimates of anticancer drug efficacy that is disentangled from TGI due to depletion of host energy. These findings strongly suggest the adoption of the tumor-in-host approach in the preclinical oncological setting for a joint assessment of drug efficacy and toxicity on animal BW and for a better protocol design of the experiments. Finally, the successfully application of the DEB-approach to different host species, several anticancer agents based on different mechanisms of action and experimental settings, including combination therapies, encourages further investigations. Specific modeling efforts are focusing on taking advantages of the DEB-based paradigm as preclinical to clinical translational approach. Preliminary results are encouraging.

References:
[1] L. Carrara, S.M. Lavezzi, E. Borella, G. De Nicolao, P. Magni, I. Poggesi (2017) Current mathematical models for cancer drug discovery, Expert Opinion on Drug Discovery, 12:8, Pages 785-799
[2] S.M. Lavezzi, E. Borella, L. Carrara, G. De Nicolao, P. Magni & I. Poggesi (2018). Mathematical modeling of efficacy and safety for anticancer drugs clinical development. Expert opinion on Drug Discovery, 13(1), Pages 5-21.
[3] I. M. van Leeuwen, C. Zonneveld, S.A. Kooijman (2003). The embedded tumour: host physiology is important for the evaluation of tumour growth. British journal of cancer, 89(12), Pages 2254-2263
[4] S.A. Kooijman (1993) Dynamic energy budgets in biological systems. Cambridge University Press.
[5] M. Simeoni, P. Magni, C. Cammia, G. DeNicolao, V. Croci, E. Pesenti, M. Germani, I. Poggesi, M. Rocchetti (2004) Predictive pharmacokinetic-pharmacodynamic modeling of tumor growth kinetics in xenograft models after administrations of anticancer agents, Cancer Research, 64, Pages 1094-1101
[6] M. Rocchetti, F. Del Bene, M. Germani, F. Fiorentini, I. Poggesi, E. Pesenti, P. Magni, G. De Nicolao. Testing additivity of anticancer agents in pre-clinical studies: A PK/PD modelling approach. Europena Journal of Cancer, 45(18)
[7] N. Terranova, E.M. Tosca, E. Borella, E. Pesenti, M. Rocchetti, P. Magni, (2018) Modeling tumor growth inhibition and toxicity outcome after administration of anticancer agents in xenograft mice: A Dynamic Energy Budget (DEB) approach, Journal of Theoretical Biology, 450, Pages 1-14
 [8] E.M. Tosca, M. Rocchetti, P. Magni, (2017) A PK/PD model for tumor-in-host growth kinetics following administration of an antiangiogenic agent given alone or in combination regimens, PAGE 26, Abstr 7168
[9] E.M. Tosca, M. Rocchetti, E. Pesenti, P. Magni, (2018) Modeling resistance development to Bevacizumab in xenograft experiments by coupling hypoxia-mediated mechanism and a Dynamic Energy Budget (DEB) based tumor-in-host model., Abstract at ACopP9
[10] E.M. Tosca, M.C. Pigatto, T. Dalla Costa, P.Magni, (2019) A Population Dynamic Energy Budget-Based Tumor Growth Inhibition Model for Etoposide Effects on Wistar Rats, Pharmaceutical Research, Volume 36