Modeling T-Cell Trafficking to Increase the Likelihood of Radiation-Induced Abscopal Effects

Heiko Enderling, PhD

Abstract

The combination of radiation and immunotherapy is currently enjoying unprecedented attention as a treatment strategy for patients with metastatic cancer. Clinical case studies and proof-of-principle clinical trials report on systemic, abscopal responses to the combination of focal irradiation and immunotherapy in patients who were progressing on immunotherapy alone.¹ However, individualized treatment plans to optimally exploit the synergy between immunotherapy and radiotherapy remain elusive due to high intra- and inter-patient heterogeneity and a myriad of possible radiation fractionation protocols, immunotherapy agents, and scheduling options.

Integrated mathematical oncology provides tools that could dissect this complexity and contribute to the transition of synergistic radiation and immunotherapy and radiation-induced abscopal effects into the personalized medicine era. To this end, we need to develop tractable, quantitative models based on carefully selected cancer biology and immunology principles. Predictions of such models, calibrated with patient-specific clinical data, need to be validated in prospective clinical trials.

Introduction

As a first step toward developing quantitative models, we recently developed a mathematical framework to simulate the systemic dissemination of T cells activated in response to focal therapy.² Model simulations suggest that metastatic sites within individual patients do not participate equally in immune surveillance and thus are likely to exhibit different systemic responses after local irradiation. We hypothesized that such a model could help identify patient-specific radiation treatment targets that have a high likelihood of inducing abscopal effects. Such targeted treatment strategies would be then worthy of validation in a prospective clinical trial. In a subsequent commentary, this model was critically discussed, with a focus on complex biology that was not incorporated in the model.³ Here we discuss the raised concerns in light of the purpose and applicability of the mathematical model. We echo the need for clinical validation.

Clinical immunotherapy trials, especially if combined with adjuvant focal cytotoxic therapies, have generated encouraging results, including evidence of clinical remissions.^4,5 Radiation therapy can synergize with immunotherapies, such as anti-CTLA4 antibodies or FMS-like tyrasine kinase 3 ligands to generate systemic responses outside the radiation field, as observed in a series of seminal studies by Demaria and Formenti.^6-8 Immune activation after radiation-induced immunogenic cell death provides an explanation for the previously believed anecdotal abscopal effect. Clinical case reports of radiation-induced abscopal effects date back to the 1950s⁹ and are reported after other immune-activating local therapies such as thermotherapies, including hyperthermia, radiofrequency ablation, and cryotherapy.^10-12

The FDA approval of multiple immunotherapeutic agents generates both a clinical need and a prime opportunity to explore and exploit radiation and immunotherapy synergy, particularly for patients with metastatic cancer. In a recent proof-of-principle clinical trial combining local radiotherapy and granulocyte-macrophage colony-stimulating factor, 11 of 41 patients exhibited an objective systemic (abscopal) response.¹ However, which metastatic sites were irradiated remained a heuristic decision.

Identifying patient-specific treatment targets adds an additional layer of personalization based upon limited clinical and biological data available for decision making (Figure). Unraveling the complex, adaptive tumor—immune system interactions that determine a response to therapy—both locally in the primary tumor and systemically in metastatic disease—requires nonlinear understanding and analysis of the multifactorial dynamics that govern them. As identified by Demaria and Formenti, more basic and translational research is needed to decrease treatment outcome uncertainty associated with the biological complexity of these interactions.³ Such research should include: 1) best radiation technique and fractionation protocol to induce antitumor immunity, 2) sequencing of immune modulators, 3) radiation and immunotherapy sensitivity of different tumor types in primary or metastatic tissue environments, 4) possible lack of common expression of antigen(s) or neoantigens between the irradiated and unirradiated metastases, and 5) component of the local immune environment, such as the availability and infiltration of dendritic cells.

Figure. Road map of abscopal effect focused clinical decision-making process. Proposed utilization place for T-cell trafficking framework is highlighted in red.

As a complementary and arguably synergistic methodology to in vitro and in vivo models, mathematical models may contribute to translating basic biological principles into clinical decision making.^13-16 We argue that some of the above-listed uncertainties may be addressed with a quantitative modeling approach where actionable data are available. Tractable, quantitative models can be designed, based on fundamental principles of cancer biology and immunology, to isolate and investigate key mechanisms and relationships within inherently complex biological systems. At present, attempting to elucidate the mechanisms that govern patient response to treatment based on a vast wealth of biological knowledge, but only a minimal amount of clinically obtainable patient information, is a near-impossible task. Instead, the tools of mathematical oncology allow us to identify key players in biological processes by using readily obtainable clinical and experimental data and evaluating its role in the broader setting of patient response. Where such data-driven models demonstrate the ability to reproduce or predict actual clinically observed phenomena, the mechanism in question can be assessed in a prospective clinical trial.

Mathematical models of tumor-immune interactions and therapy¹⁷ may be informed, parameterized, and calibrated by the experimental and clinical data collected by Demaria and Formenti^7,8 and others.¹⁸ Mathematical models, like all biological modeling systems including cell culture, tissue culture, or orthotopic xenograft models, to name but a few, are simplifications of tumors in the human body. Thus, have inherent weaknesses. The only question of interest is if the models are illuminating and useful.¹⁹ Poleszczuk and colleagues recently developed a mathematical model of T-cell trafficking in the circulatory system of the human body.² Based on physiological blood flow properties and variation in metastatic tumor volumes and anatomic disease distribution, model simulations revealed that metastatic sites within a patient, when treated with localized therapy, may have different abilities to induce systemic responses. This suggests that the metastases to irradiate in each patient mark an additional, previously unappreciated confounding factor for abscopal responses in the clinic. The mathematical model could help identify tumor sites in individual patients that serve as immunogenic hubs. The model was not designed to, nor can it, predict abscopal effects; it may, however, help to identify patient-specific treatment targets that have the highest potential to mediate an abscopal effect.

In a recent commentary to Poleszczuk’s study², Demaria and Formenti criticize the simplicity of this modeling approach, as it fails to incorporate a myriad of complex biological processes.³ One major argument is that the importance of activated T-cell dissemination patterns in triggering the abscopal effect is dominated by other factors, such as the availability of dendritic cells. Although a larger relative importance of these other factors is undoubtedly possible, this statement lacks conclusive support. Without such backing, the entropy of activated T-cell distribution among metastatic sites could still be the apparent driver of biological reactions in unirradiated metastases, independent of all other factors, despite their relevance to the question at hand. We agree that the model needs clinical validation with both retrospective and prospective data to investigate the generated biological hypotheses and to determine clinical applicability. There is encouraging clinical results suggesting that anatomical distribution of metastases could impact the likelihood of response to checkpoint blockade. In a recent phase II multicenter clinical trial of atezolizumab (Tecentriq) for metastatic urothelial carcinoma, patients with liver or visceral metastases had a lower objective response rate compared with patients without.²⁰

Predicting T-cell trafficking based on biological and physiological mechanisms, while certainly only 1 of many biological processes in abscopal responses, is exclusively calculable and predictable from routinely collected noninvasive imaging, independent of the primary tumor site. It remains currently infeasible to access the mutation load or immune infiltration in all metastatic sites in order to profile T-cell repertoires and to analyze intra- and intermetastases antigen heterogeneity prior to focal radiation. Despite recent radiomics advances²¹, current radiographic assessments cannot predict immunogenic cell death following prospective irradiation, neither in a single tumor nor as a multimetastases comparison. We will extend our mathematical framework2 beyond its current simplicity when corresponding biological and clinical data become available to inform the model.

Affiliations

Disclosures

Disclosures

We applaud Demaria and Formenti for the proposal to generate an annotated registry of abscopal responders to inform, calibrate, and validate mathematical frameworks to help improve the likelihood of triggering abscopal responses. To date, however, there are too few abscopal effect studies with coherent data to create a statistically meaningful database. In addition to suggested anecdotal case reports,^22-24 an ideal patient cohort to populate such an annotated registry comprises patients accrued to a clinical trial who were selected based on unifying clinical features and received the same treatment and longitudinal analysis. It is of equal importance to include nonresponder data in such a registry, as model predictability must also be guaranteed for patients without abscopal responses. A concerted effort must be made to ensure high-value clinical trial data from accrued patients is shareable after trial conclusion. This may bring together clinicians as well as quantitative and life scientists to ultimately advance our understanding of the complex biology underlying abscopal effects and to inform the next generation of clinical trials with available tools to understand and optimally apply synergistic treatment.Jan Poleszczuk, Rachel Walker, and Heiko Enderling are with the Department of Integrated Mathematical Oncology, H. Lee Moffitt Cancer Center & Research Institute; Eduardo G. Moros and Heiko Enderling are with Department of Radiation Oncology, H. Lee Moffitt Cancer Center & Research Institute; Mayer Fishman is with the Department of Genitourinary Oncology, H. Lee Moffitt Cancer Center & Research Institute; Julie Djeu is with the Department of Immunology, H. Lee Moffitt Cancer Center & Research Institute; Jonathan D. Schoenfeld is with the Department of Radiation Oncology Department of Radiation Oncology, Brigham and Women’s Hospital and Dana-Farber Cancer Center; and Steven Finkelstein is with the NRG Immunotherapy Committee.Heiko Enderling and Jan Poleszczuk are supported by the Personalized Medicine Award 09-33504-14-05 and 09-33000-15-03 from the DeBartolo Family Personalized Medicine Institute Pilot Research Awards in Personalized Medicine (PRAPM).Conceptual design: Heiko Enderling, Eduardo G. Moros, Mayer Fishman, Julie Djeu, Jonathan D. Schoenfeld, Steven Finkelstein, Jan Poleszczuk. Manuscript writing and review: Heiko Enderling, Jan Poleszczuk, and Rachel Walker. All authors read and approved the final manuscript.

Send correspondence to: Heiko Enderling, PhD, Moffitt Cancer Center, 12902 USF Magnolia Drive, Tampa, FL 33612; e-mail: heiko.enderling@moffitt.org.

References

1. Golden EB, Chhabra A, Chachoua A, Adams S, Donach M, Fenton-Kerimian M, et al. Local radiotherapy and granulocyte-macrophage colony-stimulating factor to generate abscopal responses in patients with metastatic solid tumours: a proof-of-principle trial. Lancet Oncol. 2015 Jul;16(7):795—803.
2. Poleszczuk J, Luddy KA, Prokopiou S, Robertson-Tessi M, Moros EG, Fishman M, et al. Abscopal benefits of localized radiotherapy depend on activated T cell trafficking and distribution between metastatic lesions. Cancer Res. 2016;76(5):1009—18.
3. Demaria S, Formenti SC. Can abscopal effects of local radiotherapy be predicted by modeling T cell trafficking? Journal for ImmunoTherapy of Cancer. BioMed Central; 2016;4(1):29.
4. Finkelstein SE, Iclozan C, Bui MM, Cotter MJ, Ramakrishnan R, Ahmed J, et al. Combination of external beam radiotherapy (EBRT) with intratumoral injection of dendritic cells as neo-adjuvant treatment of high-risk soft tissue sarcoma patients. Int. J. Radiat. Oncol. Biol. Phys. 2012 Feb 1;82(2):924—32. PMCID: PMC4241354.
5. Crittenden M, Kohrt H, Levy R, Jones J, Camphausen K, Dicker A, et al. Current Clinical Trials Testing Combinations of Immunotherapy and Radiation. Semin Radiat Oncol. 2015 Jan;25(1):54—64. PMCID: PMC4640687.
6. Pilones KA, Vanpouille-Box C, Demaria S. Combination of Radiotherapy and Immune Checkpoint Inhibitors. Semin Radiat Oncol. 2015 Jan;25(1):28—33.
7. Demaria S, Ng B, Devitt ML, Babb JS, Kawashima N, Liebes L, et al. Ionizing radiation inhibition of distant untreated tumors (abscopal effect) is immune mediated. Int. J. Radiat. Oncol. Biol. Phys. 2004 Mar 1;58(3):862—70.
8. Dewan MZ, Galloway AE, Kawashima N, Dewyngaert JK, Babb JS, Formenti SC, et al. Fractionated but not single-dose radiotherapy induces an immune-mediated abscopal effect when combined with anti-CTLA-4 antibody. Clin. Cancer Res. 2009 Sep 1;15(17):5379—88. PMCID: PMC2746048.
9. Abuodeh Y, Venkat P, Kim S. Systematic review of case reports on the abscopal effect. Current Problems in Cancer. 2016 Jan;40(1):25—37.
10. Toraya-Brown S, Fiering S. Local tumour hyperthermia as immunotherapy for metastatic cancer. Int J Hyperthermia. Taylor & Francis; 2014 Dec;30(8):531—9. PMCID: PMC4558619.
11. O'Brien MA, Power DG, Clover AJP, Bird B, Soden DM, Forde PF. Local tumour ablative therapies: opportunities for maximising immune engagement and activation. Biochim. Biophys. Acta. 2014 Dec;1846(2):510—23.
12. Shafirstein G, Kaufmann Y, Hennings L, Siegel E, Griffin RJ, Novák P, et al. Conductive interstitial thermal therapy (CITT) inhibits recurrence and metastasis in rabbit VX2 carcinoma model. Int J Hyperthermia. 2009;25(6):446—54. PMCID: PMC2885861.
13. Anderson ARA, Quaranta V. Integrative mathematical oncology. Nat. Rev. Cancer. 2008 Mar;8(3):227—34.
14. Rockne RC, Rockhill JK, Mrugala MM, Spence AM, Kalet I, Hendrickson K, et al. Predicting the efficacy of radiotherapy in individual glioblastoma patients in vivo: a mathematical modeling approach. Phys Med Biol. IOP Publishing; 2010 Jun 21;55(12):3271—85. PMCID: PMC3786554.
15. Yankeelov TE, Atuegwu N, Hormuth D, Weis JA, Barnes SL, Miga MI, et al. Clinically relevant modeling of tumor growth and treatment response. Sci Transl Med. 2013 May 29;5(187):187ps9—187ps9. PMCID: PMC3938952.
16. Prokopiou S, Moros EG, Poleszczuk J, Caudell J, Torres-Roca JF, Latifi K, et al. A proliferation saturation index to predict radiation response and personalize radiotherapy fractionation. Radiat Oncol. 2015;10(1):159. PMCID: PMC4521490.
17. Walker R, Enderling H. From concept to clinic: Mathematically informed immunotherapy. Current Problems in Cancer. 2016 Jan;40(1):68—83.
18. Chandra RA, Wilhite TJ, Balboni TA, Alexander BM, Spektor A, Ott PA, et al. A systematic evaluation of abscopal responses following radiotherapy in patients with metastatic melanoma treated with ipilimumab. Oncoimmunology. 2015 Nov;4(11):e1046028. PMCID: PMC4589983.
19. Box GEP. Robustness in the Strategy of Scientific Model Building. Robustness in Statistics. Elsevier; 1979. p. 201—36.
20. Gillies RJ, Kinahan PE, Hricak H. Radiomics: Images Are More than Pictures, They Are Data. Radiology. 2016 Feb;278(2):563—77. PMCID: PMC4734157.
21. Postow MA, Callahan MK, Barker CA, Yamada Y, Yuan J, Kitano S, et al. Immunologic correlates of the abscopal effect in a patient with melanoma. N. Engl. J. Med. 2012 Mar 8;366(10):925—31. PMCID: PMC3345206.
22. Golden EB, Demaria S, Schiff PB, Chachoua A, Formenti SC. An abscopal response to radiation and ipilimumab in a patient with metastatic non-small cell lung cancer. Cancer Immunol Res. 2013 Dec;1(6):365—72. PMCID: PMC3930458.
23. Schoenfeld JD, Mahadevan A, Floyd SR, Dyer MA, Catalano PJ, Alexander BM, et al. Ipilmumab and cranial radiation in metastatic melanoma patients: a case series and review. Journal for ImmunoTherapy of Cancer. BioMed Central; 2015;3(1):50. PMCID: PMC4678639.