Jannet Beukema

THE RELEVANCE OF CARDIAC TOXICITY IN RADIOTHERAPY FOR ESOPHAGEAL CANCER Jannet C. Beukema

THE RELEVANCE OF CARDIAC TOXICITY IN RADIOTHERAPY FOR ESOPHAGEAL CANCER Jannet C. Beukema

Copyright 2024 © Jannet Beukema All rights reserved. No parts of this thesis may be reproduced, stored in a retrieval system or transmitted in any form or by any means without permission of the author. Provided by thesis specialist Ridderprint, ridderprint.nl Printing: Ridderprint Layout and design:Yasmine Medjadji, persoonlijkproefschrift.nl

Table of Contents Chapter 1 Introduction 6 Chapter 2 Review: Is cardiac toxicity a relevant issue in the radiation treatment of esophageal cancer? 16 Published in Radiotherapy and Oncology 2015, Volume 114(85-90) Chapter 3 Editorial: Cardiac toxicity in the radiation treatment of esophageal cancer: an emerging concern. 32 Published in Future Cardiology 2015 Volume 11(4), 367-369 Chapter 4 Can we safely reduce the radiation dose to the heart while compromising the dose to the lungs in esophageal cancer patients? 38 Published in Radiotherapy and Oncology 2020, Volume 149(222-227) Chapter 5 Radiation induced myocardial fibrosis in long-term esophageal cancer survivors. 58 Published in Int J Radiation Oncol Biol Phys 2021, Volume 110(4), 1013-1021 Chapter 6 Late cardiac toxicity of neo-adjuvant chemoradiation in esophageal cancer survivors: a cross-sectional pilot study. 78 Published in Radiotherapy and Oncology 2022, Volume 167(72-77) Chapter 7 Blood biomarkers for cardiac damage during and after radiotherapy for esophageal cancer: a prospective cohort study 96 Published in Radiotherapy and Oncology 2024, Volume 200 Chapter 8 Summarized discussion and future perspectives. 114 Appendices Nederlandse samenvatting 134 Access to supplementary data 140 Curriculum Vitae 141 List of publications 142 Dankwoord 146

CHAPTER 1 General introduction

8 Chapter 1 Introduction In the Netherlands, approximately 3000 new esophageal cancer patients are diagnosed each year and its incidence is still rising, which is especially true for adenocarcinomas located in the distal part of the esophagus. About 60% of patients present with potentially curable disease and the majority undergoes neoadjuvant chemoradiotherapy followed by surgery or definitive (chemo)radiotherapy[1]. Although cure rates improved over the last decade, treatment-induced toxicity is still a matter of concern[2]. For most esophageal cancer patients, radiotherapy target volumes are relatively large and located near critical organs like the heart and the lungs. Therefore, high toxicity rates related to these organs at risk can be expected. However, trade-offs between cardiac and pulmonary toxicities, and thus decisions on how to optimize dose distributions in radiotherapy treatment planning, require more detailed information on toxicities in relation to radiation dose distributions. At the time of the start of this thesis, literature from clinical trials mainly focused on pulmonary toxicity and its relationship with lung dose volume parameters. Whereas literature on radiation induced cardiac toxicity, specifically in esophageal cancer patients, was scarce (Figure 1). For this reason, we decided to focus this thesis on radiation-induced cardiac toxicity in the treatment of esophageal cancer patients. Table 1 timeline of PubMed results when comparing the number of clinical trials on radiation AND “cardiac toxicity” vs “pulmonary toxicity”.

9 General introduction Newer techniques in radiotherapy The last decades, radiation technologies have been significantly improved. In the 90’s, three-dimensional conformal radiotherapy (3D-CRT) was commonly used. More recently, more advanced technologies like intensity modulated radiotherapy (IMRT) and volumetric modulated arc therapy (VMAT) became the new standard. These technologies can deliver highly conformal dose distributions with improved ratios between target coverage and sparing of critical OARs, like the heart and lungs (Figure 1). Moreover, these techniques provide additional flexibility in prioritizing which OARs should be avoided. This prioritizing of dose to critical organs remains a key issue as, especially in photon radiotherapy, as decreasing the dose to one organ, for example the heart, will come at the expense of the radiation dose to other organs, like the lungs. The new kid on the block, proton radiotherapy is an even more advanced technology, which allows a further reduction of the radiation dose to both heart and lungs. However, its availability is limited and therefore it is important to select the patients that benefit most[3]. Figure 1, evolution of radiotherapy planning techniques in recent years, techniques do become more conformal Red=target volume, Green=prescribed dose, Blue=intermediate dose, 40-80% prescribed dose, Grey=low dose, up to 20% prescribed dose. 1

10 Chapter 1 Dose volume histograms A dose volume histogram (DVH) in radiotherapy represents the 3 dimensional dose distribution of target volumes or critical organs in a 2D graph. This DVH graph is a cumulative graph, with the radiation dose (in Gray) on the x-axis and the volume (in percentages) on the Y axis. These DVH graphs can be used to compare different treatment plans or techniques. DVH parameters can be extracted from these graphs and are used in scientific literature and radiotherapy guidelines. Values like D98(Gy) are used to describe the plan quality (target coverage) and represents the minimum dose given to 98% of the target volume. V values are used to describe the dose on OARs. E.g., V20 of the heart represents the volume of the heart in percentage that receives 20 Gy or more. Normal tissue complication probability Normal Tissue Complication Probability (NTCP) models describe the relation between radiation dose distributions in one or more OARs and the risk of complications. These prediction models are developed by using cohorts of patients who were treated in the past. Next to Dose Volume Histogram (DVH) parameters, other clinical risk factors (like age, smoking, or a cardiac history) can be included in these models to correct for confounding or to improve model performance. These multivariable NTCP-models should preferably be based on large patient cohorts, and validated in independent patient cohorts to assess generalizability in other study populations. These externally validated multivariable prediction models are currently considered the highest level of evidence for the prediction of treatment related complications [4]. Radiation induced pulmonary toxicity. As mentioned before, most literature in the 90’s focused on pulmonary toxicity. Radiation-induced pulmonary toxicity may present with different clinical symptoms, varying from mild dyspnea and non-productive cough to respiratory failure requiring mechanical ventilation which could eventually be fatal. Currently, the most widely used NTCP-models for radiation-induced pulmonary toxicity include the mean lung dose as DVH parameter[5,6], next to clinical factors like age, co-morbidities and the location of the tumor [7].

11 General introduction Radiation induced cardiac toxicity. Most clinical publications on radiation-induced cardiac toxicity were published after 2015. Historically, cardiac toxicity was considered a (very) late event and increased rates of cardiovascular diseases were observed in long term survivors of breast cancer and Hodgkin lymphoma[8,9]. Incidence rates of myocardial infarction (HR: 1.22(95%CI:1.06-1.42)), pericarditis (HR: 1.61(95%CI: 1.06-2.43)) as well as valve disorders (HR: 1.54(95%CI: 1.11-2.13)) were higher in left-sided breast cancer patients that were treated with radiotherapy as compared to right-sided breast cancer patients. For Hodgkin lymphoma survivors, an increased rate of heart failure (4.9 times) as well as myocardial infarctions (3.6 times) has been observed as compared to normal populations. The combination of radiotherapy and chemotherapy (anthracyclines) resulted in the highest risk of late cardiac toxicity in this population[10]. Literature on cardiac toxicity and radiation dose parameters is limited. The only externally validated NTCP model originates from breast cancer patients. In 2013, Darby et al published this data. The rate of major coronary events increased linearly with the mean radiation dose to the heart by 7.4 % per Gray[11]. These data were validated in another breast cancer population by Boogaard et al [12]. Overall survival as a surrogate for toxicity endpoints? As causes of death are often unknown, overall survival (OS) and/or death have been used as alternative endpoints in relation to normal tissue dose. This probably originated from trials were better tumor specific survival or local control was observed, while OS rates did not improve or even worsened[13,14]. Analyzing OS as an endpoint instead of toxicity is attractive because it can combine toxicities of, for example lung and heart, while considering tumor prognostic factors, like tumor stage. Moreover, it is a truly relevant endpoint. However, OS does not provide information on causal relationships. In addition, OS is less sensitive and provides less guidance on which specific regions of the thoracic region you should try to spare in radiotherapy treatment planning. In the treatment of intrathoracic tumors, several prediction models for OS have been developed [5,15,16]. These models include tumor-specific prognostic factors like tumor size or lymph node status next to DVH-parameters of the heart. Only Speirs et included a lung DVH parameter next to a heart DVH parameter in their final multivariable prediction model. The model of Defraene et al. included tumor 1

12 Chapter 1 size next to smoking (a protective factor for OS) and the mean dose to the heart in their prediction model. This latter model has been validated in another lung cancer cohort as well as in an esophageal cancer cohort[17]. These externally validated prediction models are currently used for optimizing radiotherapy dose distributions and selecting patients for newer techniques like proton therapy in the Netherlands. Prediction models can be more reliable and reproducible, when you know more about the pathogenesis. This knowledge may help in the preselection of the DVH parameters of (subregions of) critical OARs while developing prediction models [15]. Besides, this may guide choosing the most relevant DVH-parameters in case of multicollinearity, as different anatomically closely related DVH parameters, like the heart, its subregions, and the lungs often have a strong correlation with each other. Summarizing, although prediction models suggest that irradiation of the heart leads to worse OS, a causal relationship between radiation dose to the heart and toxicity cannot be concluded from these papers. Based on the limited evidence available on radiation induced cardiac toxicity in esophageal cancer, we started this thesis to improve knowledge on the relevance and mechanisms of radiation induced cardiac toxicity as a first step towards NTCP modelling and to guide trade-offs between radiation dose to the heart and the lungs during the treatment planning process. Outline of this thesis Chapter 1 is the introduction of this thesis, in which provided background information on the literature at start of the thesis. Furthermore, we explain some of the terminology used and the end we highlight the unmet needs on cardiac toxicity in the treatment of esophageal cancer. In chapter 2 we performed a review of the available literature on radiation induced cardiac toxicity in the treatment of esophageal cancer. Chapter 3 consists of an editorial on radiation induced cardiac toxicity to increase awareness under other medical disciplines. Chapter 4 describes retrospective cohort study based on a population of 216 esophageal cancer patients who underwent curative radio(chemo)therapy to a relatively high radiation dose. We performed multivariable analyses combining clinical data with dose volume parameters to develop multivariable prediction models for heart as well as lung toxicity.

13 General introduction Chapter 5 and 6 describe the results of a cross-sectional prospective study among esophageal cancer survivors treated with neoadjuvant chemoradiotherapy plus surgery versus patients treated with surgery only. We performed several diagnostic tests to objectify clinical as well as subclinical damage that might be caused by the chemoradiation in these so-called survivors. In chapter 5 we focused more specifically on changes seen on MRI, quantified these findings and related them to the spatial dose distribution as given during the neoadjuvant chemoradiotherapy. In chapter 6 we combined the results of the different imaging techniques, ECG, biomarkers and functional assessments versus clinical events and dose distributions. Chapter 7 reports on a longitudinal prospective study monitoring cardiac blood biomarkers during and after radiotherapy for esophageal cancer. We hypothesized that these parameters would be useful in future clinical practice and/or trials to predict radiation induced cardiac toxicity. We analyzed relationships with cardiac and pulmonary radiation dose volume parameters and clinical events of both organs at risk. In chapter 8 we conclude by summarizing the results and discuss future perspectives 1

14 Chapter 1 References [1] Integraal Kankercentrum Nederland. Slokdarm- en maagkanker in Nederland 2021:1–48. [2] van Hagen P, Hulshof MC, van Lanschot JJ, Steyerberg EW, van Berge Henegouwen MI, Wijnhoven BP, et al. Preoperative chemoradiotherapy for esophageal or junctional cancer. N Engl J Med 2012;366:2074–84. https://doi.org/10.1056/NEJMoa1112088. [3] Langendijk J a, Lambin P, De Ruysscher D, Widder J, Bos M, Verheij M. Selection of patients for radiotherapy with protons aiming at reduction of side effects: the model-based approach. Radiother Oncol 2013;107:267–73. https://doi.org/10.1016/j.radonc.2013.05.007. [4] Moons KGM, Altman DG, Reitsma JB, Ioannidis JPA, Macaskill P, Steyenberg EW, et al. Transparent Reporting of a multivariable prediction model for Individual Prognosis or Disagnosis (TRIPOD): Explanantion and Elaboration. Ann Intern Med 2015;162:W1–74. https://doi.org/10.7326/M140698. [5] Defraene G, Schuit E, De Ruysscher D. Development and internal validation of a multinomial NTCP model for the severity of acute dyspnea after radiotherapy for lung cancer. Radiother Oncol 2019;136:176–84. https://doi.org/10.1016/j.radonc.2019.03.034. [6] KWA SLS. JVL. Radiation pneumonitis as a function of mean lung dose: an analysis of pooled data of 540 patients. Int J Radiat Oncol Biol Phys 1998;42:1–9. [7] Vogelius IR, Bentzen SM. A literature-based meta-analysis of clinical risk factors for development of radiation induced pneumonitis. Acta Oncol (Madr) 2012;51:975–83. https://doi.org/10.3109/0 284186X.2012.718093. [8] Aleman BMP, van den Belt-Dusebout AW, Klokman WJ, Van’t Veer MB, Bartelink H, van Leeuwen FE. Long-term cause-specific mortality of patients treated for Hodgkin’s disease. J Clin Oncol 2003;21:3431–9. https://doi.org/10.1200/JCO.2003.07.131. [9] McGale P, Darby SC, Hall P, Adolfsson J, Bengtsson N-O, Bennet AM, et al. Incidence of heart disease in 35,000 women treated with radiotherapy for breast cancer in Denmark and Sweden. Radiother Oncol 2011;100:167–75. https://doi.org/10.1016/j.radonc.2011.06.016. [10] Aleman BMP, Van Den Belt-Dusebout AW, De Bruin ML, Van ’t Veer MB, Baaijens MHA, De Boer JP, et al. Late cardiotoxicity after treatment for Hodgkin lymphoma. Blood 2007;109:1878–86. https://doi.org/10.1182/blood-2006-07-034405. [11] Darby SC, Ewertz M, McGale P, Bennet AM, Blom-Goldman U, Brønnum D, et al. Risk of ischemic heart disease in women after radiotherapy for breast cancer. N Engl J Med 2013;368:987–98. https://doi.org/10.1056/NEJMoa1209825. [12] Van Den Bogaard VAB, Ta BDP, Van Der Schaaf A, Bouma AB, Middag AMH, Bantema-Joppe EJ, et al. Validation and modification of a prediction model for acute cardiac events in patients with breast cancer treated with radiotherapy based on three-dimensional dose distributions to cardiac substructures. J Clin Oncol 2017;35:1171–8. https://doi.org/10.1200/JCO.2016.69.8480. [13] “Port Meta-analysis Trialists group.” Postoperative radiotherapy in non-small-cell lung cancer: systematic review and meta-analysis of individual patient data from nine randomised controlled trials. PORT Meta-analysis Trialists Group. Lancet 1998;352:257–63. [14] Bradley JD, Paulus R, Komaki R, Masters G, Blumenschein G, Schild S, et al. Standard-dose versus high-dose conformal radiotherapy with concurrent and consolidation carboplatin plus paclitaxel with or without cetuximab for patients with stage IIIA or IIIB non-small-cell lung cancer (RTOG 0617): a randomised, two-by-two factorial p. Lancet Oncol 2015;16:187–99. https://doi. org/10.1016/S1470-2045(14)71207-0. [15] Speirs CK, DeWees TA, Rehman S, Molotievschi A, Velez MA, Mullen D, et al. Heart Dose Is an Independent Dosimetric Predictor of Overall Survival in Locally Advanced Non–Small Cell Lung Cancer. J Thorac Oncol 2017;12:293–301. https://doi.org/10.1016/j.jtho.2016.09.134. [16] Atkins KM, Rawal B, Chaunzwa TL, Lamba N, Bitterman DS, Williams CL, et al. Cardiac Radiation Dose, Cardiac Disease, and Mortality in Patients With Lung Cancer. J Am Coll Cardiol 2019;73:2976–87. https://doi.org/10.1016/j.jacc.2019.03.500.

15 General introduction [17] Berbée M, Muijs CT, Voncken FEM, Wee LS, Sosef M, van Etten B, et al. External validation of a lung cancer-based prediction model for two-year mortality in esophageal cancer patient cohorts. Radiother Oncol 2024;190. https://doi.org/10.1016/j.radonc.2023.109979. 1

CHAPTER 2 Is cardiac toxicity a relevant issue in the radiation treatment of esophageal cancer? Jannet C. Beukema, MD1, Peter van Luijk, PhD1, Joachim Widder, MD, PhD1, Johannes A. Langendijk, MD, PhD1, Christina T. Muijs, MD, PhD1 1Department of Radiation oncology, University Medical Center Groningen, Groningen, The Netherlands. Published in Radiotherapy and Oncology Volume 114(1), 2015, 85-90

18 Chapter 2 Abstract Purpose: In recent years several papers have been published on radiation-induced cardiac toxicity, especially in breast cancer patients. However, in esophageal cancer patients the radiation dose to the heart is usually markedly higher. To determine whether radiation-induced cardiac toxicity is also a relevant issue for this group, we conducted a review of the current literature. Methods: A literature search was performed in MEDLINE for papers concerning cardiac toxicity in esophageal cancer patients treated with radiotherapy with or without chemotherapy. Results: The overall crude incidence of symptomatic cardiac toxicity was as high as 10.8 %. Toxicities corresponded with several dose-volume parameters of the heart. The most frequently reported complications were pericardial effusion, ischemic heart disease and heart failure. Conclusion: Cardiac toxicity is a relevant issue in the treatment of esophageal cancer. However, valid Normal Tissue Complication Probability models for esophageal cancer are not available at present.

19 Review on cardiac toxicity Introduction Increasing numbers of patients with esophageal cancer are currently being treated with curatively intended combined modality strategies such as chemoradiation, either in the neoadjuvant setting followed by surgery or as definitive treatment. Due to improved outcome, resulting from the addition of chemotherapy to radiotherapy, more patients are surviving treatment of esophageal cancer. Therefore, more patients are at risk for treatment-related toxicity, which is becoming a major concern. Little data is available on cardiac morbidity and mortality following radiation treatment of esophageal cancer. This may be explained by the fact that cardiac toxicity has traditionally been regarded as a late (or very late) side effect. Given the relatively low incidence of esophageal cancer and previously low cure rates after treatment, data on radiation-induced cardiac toxicity has remained scarce. Radiation-induced cardiac toxicity in other cancers, such as breast cancer and Hodgkin’s lymphoma, has been studied more extensively due to larger numbers of patients and better survival. In these patient groups, higher rates of cardiac death have been found during long-term follow up [1]. However, in esophageal cancer, cardiac doses are generally markedly higher due to the location of the target area close to the heart and/or to the higher total dose. Currently, no clear guidelines exist on how to distribute the radiation dose between the different organs at risk in radiotherapy treatment planning of esophageal cancer, in particular the relation between the dose to the lungs and to the heart. This question becomes even more relevant with the clinical introduction of new radiation delivery techniques like intensity modulated radiotherapy (IMRT) and proton therapy. We therefore reviewed the current evidence on types and incidence of radiation induced cardiac toxicity and the possibly association with dose-volume parameters after multimodality treatment for esophageal cancer. 2

20 Chapter 2 Methods Medline was searched for “heart”[MESH] AND “radiation therapy [MESH]” AND “ esophageal cancer [MESH]”, to retrieve papers that published on radiation induced cardiac toxicity and/or radiation dose parameters. Papers publishing data on cardiac toxicity and radiation dose parameters between 1970 and the first of July 2013 were included in this review. References of the articles were screened for other papers and included in this review when considered relevant. Figure 1

21 Review on cardiac toxicity Results The literature search resulted in a total of 38 papers, of which the abstracts were screened first for their relevance to our review. After initial screening, seven papers were considered relevant. After screening the references from these papers, another 6 papers were retrieved and included. An overview of the selection process is shown in figure 1. All selected papers are listed and briefly summarized in Table 1. Table 1. Selected papers on cardiotoxicity in chemoradiotherapy for esophageal cancer.a (Continued) Author N Total dose RT (dose/fr) (Gy) FU (months) Time to event (months) Toxicities (N) Association with dose distribution parameters Morota et al. (2009) 69 60 (2) 26.1 10 6% > grade II: pericardial effusion (n = 1), valve replacement → heart failure (n = 1), cardiac ischemia (n = 1), and pleural effusion (n = 11) Not available Ishikura et al. (2003) 139 60 (2) 53 14 11% > grade II: myocardial infarction → death (n = 2) pericardial effusion (n = 8) → grade V heart failure (n = 2), pleural effusion (n = 8) Not available Kumekawa et al. (2006) 81 60 (2) 57 Mean within 24 11% > grade II: pericardial effusion (n = 3) → grade V heart failure (n = 2), cardiac ischemia (n = 3) → grade V (n = 1), pleural effusion (n = 3) → grade V (in combination with pneumonitis) (n = 1) Not available Martel et al. (1998) 57 37.5–49 (1.5–3.5) 19 8 5% > grade II: pericardial effusion (n = 3) → grade V (n = 1) Mean and max heart dose Wei et al. (2008) 101 45–50.4 (1.8–2) 8.4 5.3 28% any pericardial effusion Pericardial dose > 26.1 Gy; V5–45 pericard Shirai et al. (2011) 43 52–70 (1.8–2) 26.9 4 35% any pleural effusion, hypertension (n = 11), arrhythmia (n = 5), ischemia (n = 2), cardiomyopathy (2), mitral regurgitation (1) Older age and V50 heart Mukherjee et al. (2003) 15 45–50 (1.8–2) Not relevant 1 80% any drop in ejection fraction 1 month after CRT No correlation with heart dose Table 1. Selected papers on cardiotoxicity in chemoradiotherapy for esophageal cancer.a 2

22 Chapter 2 Table 1. Selected papers on cardiotoxicity in chemoradiotherapy for esophageal cancer.a (Continued) Author N Total dose RT (dose/fr) (Gy) FU (months) Time to event (months) Toxicities (N) Association with dose distribution parameters Tripp et al (2005) 20 45–54 (1.8–2) Not relevant 1.5 55% any drop in ejection fraction but in 30% a rise in ejection fraction No correlation with heart dose Gayed et al. (2006) 51 50.4–60 (1.8–2) Not relevant 3 54% perfusion abnormalities and 42% inferior wall ischemia Irradiated patients; higher dose areas >45 Gy Gayedet al. (2009) 16 30–50.4 (1.8–2) 14.6 12 43% any cardiac complications: ischemia (n = 1), atrial fibrillation (n = 2), Pericardial effusion (n = 2), heart failure (n = 2) → complete heart block grade V (n = 1) Not available Konski et al. (2012) 102 45–57.6 (1.8) 10.7 4.2 12% > grade 2: pericardial effusion (n = 10), myocardial infarction (n = 1), sick sinus syndrome (n = 1) Correlation V20, V30 and V40 heart with symptomatic heart toxicity Jingu et al. (2006) 64 30–70 (2) Not relevant 9.3 20% increased uptake on FDG PET Higher SUV values within the radiation fields Hatakenaka et al. (2012) 31 41–60 (1.8–2) Not relevant 3 days Lower left ventricular end diastolic volume and stroke index, an increased heart rate and left ventricular wall motion disorders after treatment Significant difference in high vs. low left ventricular dose groups Only the paper published by Jingu et al. had a prospective design, the others were retrospective. All papers combined radiotherapy with chemotherapy, the most frequently used schedule was 5-FU and cisplatinum. Almost all patients were treated with a 3D CRT technique, IMRT was used in a few patients. Three papers reported specifically on retrospective follow-up data and late cardiopulmonary RTOG rated toxicity in esophageal cancer patients.[2–4] All patients in these studies were treated with concurrent chemoradiation to a total dose of 60 Gy in combination with cisplatin and 5-FU. Target definition and radiotherapy planning were performed with 2-dimensional techniques using simulation films. Therefore, individual cardiac dose distributions were not available and no attempts were made to correlate cardiac dose to toxicity. Patient numbers and details on

23 Review on cardiac toxicity grade 3 or higher reported cardiac toxicity in these papers are listed in and marked as the first three papers in table 1. The most frequently observed side effects were cardiac ischemia, pleural and pericardial effusions and heart failure. These “late toxic events” presented occurred relatively soon after treatment, with a median follow up of 26.1 to 57 months. Grade 3 or higher cardiac toxicity, which is considered clinically relevant, was seen in 5.8%- 11.1% of the patients. Given the low cure rates of esophageal cancer in these studies, with 3 years survival rates varying between 22% and 45%, the actuarial rates were not reported, but can be expected to be much higher. Morota et al. also reported on patient and treatment related risk factors for cardiac events.[3] Older age (>75 years) was the only factor significantly associated with late cardiopulmonary toxicity. They reported a crude incidence of 29% in the older patient group vs. 3% for the younger patients. In three other studies, the authors aimed to find clinical and dosimetric factors influencing the risk for pericardial(PCE) or pleural effusion(PE).[5–7] Treatment details are summarized as the next three papers in table 1. Martel et al. was the first to report on pericardial effusion among patients treated with 3D-CRT based on planning-CT and available diagnostic data.[5] Between 1985 and 1991, patients were treated according to 3 different protocols. The only prognostic factor significantly associated with PCE was the dose per fraction (3.5 Gy). After correction, according to the Linear Quadratic model, the mean and maximum heart doses were significant prognostic factors. However, given the relatively small sample size, fitting the data into the Lyman model showed large confidence intervals. Wei et al. performed a retrospective analysis to identify clinical and dosimetric prognostic factors for PCE in 101 patients with inoperable esophageal cancer.[6] The pericardium was contoured as a shell, by extending the actual heart contour with 0.5 cm. PCE was scored using CT scans routinely made during follow-up visits. The mean time to onset of PCE was 5.3 months, leveling off at 16.7 months after treatment. The crude incidence of PCE was 27.7% and the actuarial incidence at 18 months was 48%. No patient or treatment-related factors could be found that were associated with PCE. However, significant associations were found with several dosimetric factors. Pericardial DVH values correlated better with the incidence of PCE as compared to the cardiac DVH parameters. If the mean pericardium dose was 2

24 Chapter 2 reduced below 26.1 Gy, the risk for PCE decreased from 73% to 13% at 18 months after treatment. The strongest prognostic factor was a V30 pericardium of >46%. Shirai et al. retrospectively analyzed 43 esophageal cancer patients.[7] In total, 35% of the patients developed non-malignant PE, including 4 patients (13%) with grade ≥2, which required medical intervention. In the univariate analysis, most cardiac and one lung parameter (V50 lung) were significantly associated with the development of PE. In de multivariate analysis, older age and the cardiac V50 were the only significant prognostic factors for PE. The left ventricular ejection fraction was studied in two papers [8,9], with relatively low patient numbers. Treatment details are listed in table 1. A small decline(4-5%) in ejection fraction was found after treatment in both papers. However, no significant association was found between dose distribution and a reduction in ejection fraction. 3D functional cardiac imaging was used to evaluate cardiac toxicity in five papers. [10–14] Treatment details and patient numbers are again listed and marked as the last five papers in table 1. Gayed et al. compared 26 irradiated to 25 non-irradiated esophageal cancer patients from a prospective database[11]. Cardiac risk factors, including demographics were comparable in the two groups. In this cohort, gated myocardial perfusion scans were routinely performed preoperatively and blinded for former treatment. Perfusion abnormalities and wall ischemia were increasingly seen in the irradiated group, but functional parameters (left ventricular ejection fraction, end diastolic and systolic) did not differ significantly. Most perfusion defects were found in the higher dose areas (>45 Gy), 70% vs. 25%. However, the mean heart dose was not statistically higher in the patients with abnormal perfusion scans. In another study from the same investigators, the clinical implications of these perfusion abnormalities in 24 lung and 16 esophageal cancer patients were investigated [10]. Although new perfusion defects were seen in about 1/3 of the patients, no significant relationship was found with symptomatic cardiac complications after a rather short median follow up of 10.9 months. Konski et al. evaluated 74 esophageal cancer patients using FDG-PET.[12] The FDG uptake declined, especially in the lateral myocardial wall, shortly after treatment. A significant association was found between the V20, V30 and V40 of the heart and symptomatic cardiac toxicity. The V40 was 69.2% vs. 53.8% among patients with

25 Review on cardiac toxicity or without symptomatic cardiac toxicity, respectively. No associations were found between cardiac toxicity and medical history, surgery or decreased FDG uptake of the myocardium. In contrast, Jingu et al. found increased FDG uptake within the irradiated field after a median time of 9.3 months after treatment for esophageal cancer. In a prospective study, 8 patients underwent additional ultrasound, MRI and myocardial SPECT to investigate the state of metabolism and vascular flow in the myocardium. The SPECT studies suggested microvascular damage and impairment in perfusion and fatty acid metabolism; under these ischemic conditions, glucose metabolism increases. MRI scans with gadolinium showed delayed enhancement in only 2 patients and was thought to be relatively insensitive to myocardial damage. Hatahenaka et al. reported on the results obtained in 31 patients treated with chemoradiation.[13] Patients were subjected to cardiac MRIs before, during and shortly after therapy. Patients were divided into a low left ventricle dose group (mean LV dose of 0.33 Gy, predominantly upper and middle esophageal tumors), and a high dose group with a mean dose of 18.1 Gy. The LV ejection fraction (LVEF), LV end diastolic volume index and left ventricular stroke index were significantly lower after treatment, which was also the case for wall motion disorders in segments 8, 9 and 10. The heart rate was significantly higher after treatment. In the low dose group, only LVEF was decreased, suggesting a role for cisplatinum. In conclusion, these five imaging studies showed early wall motion disorders and reduced or increased uptake within the irradiated area on different imaging modalities, which indicates a local effect in the myocardium. Changes in the metabolism of the irradiated areas may explain these effects. Discussion This review was undertaken to evaluate the current evidence on the types and incidence of radiation induced cardiac toxicity after multimodality treatment for esophageal cancer, in order to improve radiotherapy treatment decision making. The incidence of clinically relevant cardiac complications was reported in 6 out of 10 reviewed papers. The overall crude incidence was 10.8 % (range: 5%-44%). Most events occurred within 2 years after treatment. Given the low overall survival rate of 3 years, the actuarial incidence rate for cardiac complications is expected to be much higher. 2

26 Chapter 2 The high complication rates in these retrospective studies were not confirmed by prospective (randomized) esophageal cancer trials. This could be explained by the relatively simple radiation delivery techniques used in most of the retrospective studies. Moreover, the total dose was relatively high, 60 Gy in 3 out of the 6 papers, versus 50.4 Gy which is currently the standard in many European countries and the USA. On the other hand, it is very likely that cardiac morbidity was poorly reported in the older trials because the relationship with the given treatment was not well acknowledged at that time. Studies reporting on randomized preoperative CRT do not provide much additional information regarding toxicity. Most of these trials had relatively low patient numbers . In meta-analyses, only postoperative morbidity, mortality and overall survival have been reported.[15–18]. Furthermore, the two largest trials , have a relative short follow up and one of them has not even been fully published. [18,19] Bosch et al. on the other hand focused more specifically on chemoradiation-induced morbidity. They retrospectively compared 96 patients treated with preoperative CRT (41.4 Gy/carbo/taxol) with matched controls who were treated with surgery only.[20] In this study, rates of pneumonia, cardiac arrhythmia and pleural effusion were observed more frequently in the preoperatively treated group. Despite these events, no differences in hospital stay or short-term mortality were found, which is in line with the meta-analysis data. On the other hand, in a meta-analysis investigating the role of postoperative radiotherapy for lung cancer, a relative increase in mortality of 21% was found in the irradiated group.[21] Although the authors were unable to analyze causes of death in this meta-analysis, non-cancer-related causes of death were suggested since the local recurrence rates were reduced by postoperative radiotherapy. The latter might be a relevant finding for the treatment of esophageal cancer as target volumes for lung cancer are comparable with those for esophageal cancer. As mentioned previously, it is not always clear if cardiac events are actually related to radiation treatment. A strong argument for radiation-induced cardiac toxicity in this patient group is the association with dose-volume parameters. Based on the current literature, however, it remains difficult to determine the most relevant dose parameter. An important reason is probably that the toxicities – and thus the endpoints – used in these studies were diverse. It is very unlikely that focal wall motion disorders as seen in the imaging studies correspond to the same DVH parameters as the risk of developing pericarditis.

27 Review on cardiac toxicity Not only treatment related factors were significant related to cardiac toxicity, older age and female sex gave a significantly higher risk in two[3,7] and one[12] of the reviewed papers. However, patient numbers are low and these non-treatment related risk factors should be confirmed in larger patient groups. The challenge for the future will be to decide which clinical endpoints are relevant and should be incorporated into an NTCP model for esophageal cancer patients. Pericardial effusion was the most frequently observed complication, with an actuarial rate of 48%. [6] The biological mechanism behind PE is considered inflammatory. [22] In most cases, pericardial effusion is self-limiting and asymptomatic, but may progress into heart failure and death.[2–4] The observed incidence depends heavily on the use of routine imaging techniques, such as CT scans. Although it is the easiest clinical endpoint to incorporate into a model (high rate and objective), pericarditis is not expected to have a significant impact on the quality of life of the surviving patients. Secondary ischemic events occur frequently in esophageal cancer patients treated with radiotherapy or CRT, not only as a very late side effect, but also during the first 2 years after completion of treatment. However, radiotherapy may not be the only risk factor for this cardiac event. Esophageal cancer patients generally have a number of risk factors for ischemic heart disease, including older age, histories of tobacco use and/or obesity.[23] All patients were irradiated combined with chemotherapy, most often with cisplatinum and 5-FU, which are both associated with increased risks of thrombus formation. [24–26] Based on the available data, it was impossible to correct for these potential confounding factors. Heart failure, the third most frequently observed complication after radiotherapy or chemoradiation, may be the result of other cardiac events, including myocardial infarction, pericarditis and valvular disorders. It may very well be possible that these different cardiac events with different underlying mechanisms relate to different dose-volume response relationships and as such may result in secondary cardiac events. These kind of relationships should be adjusted for in multivariable prediction models. Clinical research on functional imaging and other cardiac function parameters is necessary to better understand the mechanisms of radiation induced-cardiac toxicity and to identify the most critical parts of the heart for each of the aforementioned clinical endpoints. 2

28 Chapter 2 Most of the presented imaging studies showed wall motion disorders and changes in metabolism of the myocardium in the higher radiation dose areas. These changes are in line with results from autopsy and animal studies showing damage to microvasculature, focal ischemia and fibrosis.[22] Although these changes did not result in decreased LVEF, changes in end diastolic volume, stroke indices and heart rate were observed. The LVEF may underestimate the actual cardiac damage because of the compensatory reserve of the myocardium that enables adequate ventricular outcome even when part of the myocardium dysfunctions. Reduced end diastolic volumes are known to precede a decline of the ejection fraction.[27] As imaging studies were performed shortly after treatment, these later changes were not observed. Animal studies in mice showed progressive malfunctioning of mitochondria and progressive fibrosis in the myocardium at 40 weeks after treatment.[28,29] These changes did not result in changes in cardiac function. However, in the high dose group, a significant proportion of the mice died suddenly during follow up. These sudden deaths may imply that compensatory mechanisms may not maintain cardiac function for a longer period of time.[30] Additional clinical imaging studies at later time points after treatment as well as other cardiac functional assessments are required to get more insight into the pathophysiological mechanisms of cardiac toxicity and thus should be included in future research. Evidence regarding radiation induced cardiac toxicity is available from studies in other tumor sites with high rates of cardiac death and morbidity in irradiated patients with long term follow up.[1,31,32] Most of these articles reported frequent late or very late side effects. The studies presented in this review suggest earlier toxicity with events in the first two years after treatment. Recently, Darby et al.[1] reported on the relationship between radiation dose and major cardiac events among breast cancer patients and compared them with a population-based matched control group. They found 44% of the events in the first 10 years, with no significant trend in time. In that study, the estimated mean dose to the heart was only 4.9 Gy, which is much lower than that observed among patients treated for esophageal cancer. They observed an increase in the relative risk for ischemic heart disease of 7.4% per 1 Gy mean heart dose. There was no suggestion the increase in risk was less pronounced in the higher dose group (mean dose above 10 Gy). Interestingly, they found the mean heart dose to be more relevant for ischemic events than the dose to the left anterior descending coronary artery, but as dose reconstruction in such a small organ is less reliable as shown by Lorenzen et al, this remains to be confirmed by others.[33]

29 Review on cardiac toxicity Esophageal cancer patients are different from breast cancer and lymphoma patients, as their prognosis is poorer and the radiation doses to the heart are much higher. In current routine clinical practice, radiation oncologists consider the spinal cord and lungs as the most important critical organs. Although the results of our review confirm that the heart should also be considered to be a critical organ, the optimal distribution of the dose between the various OARs remains to be determined. In general, a reduction of the dose to the heart, even with advanced radiation delivery techniques such as IMRT or VMAT, will result in a higher lung dose with an increased risk of radiation pneumonitis and fibrosis. Proton therapy may overcome this problem, but is not widely available and relatively expensive. Therefore, selection of the patients who will benefit most from proton therapy will be essential.[34] However, if the precise association between radiation-induced side effects and the dose-volume parameters is not clear, the translation of observed differences in dose distributions between protons and photons into clinical benefits remains difficult. Accurate multivariable prediction models on radiation-induced toxicity are necessary to estimate the potential added benefits of various techniques. Although we showed that cardiac toxicity is a relevant problem in the treatment of esophageal cancer, we will obviously need more esophageal cancer patients with strict follow up data and dose distributions on critical organs as current data are insufficient to make prediction models for radiation-induced cardiac toxicity in these patients. As causes of death are often hard to identify, overall survival in addition to disease specific survival is very important to avoid underestimation of cardiac toxicity. Imaging studies and cardiac function parameters during follow up will help us identifying the most relevant clinical endpoints and critical parts of the heart. 2

30 Chapter 2 References [1] Darby SC, Ewertz M, McGale P, Bennet AM, Blom-Goldman U, Brønnum D, et al. Risk of ischemic heart disease in women after radiotherapy for breast cancer. N Engl J Med 2013;368:987–98. [2] Kumekawa Y, Kaneko K, Ito H, Kurahashi T, Konishi K, Katagiri A, et al. Late toxicity in complete response cases after definitive chemoradiotherapy for esophageal squamous cell carcinoma. J Gastroenterol 2006;41:425–32. [3] Morota M, Gomi K, Kozuka T, Chin K, Matsuura M, Oguchi M, et al. Late toxicity after definitive concurrent chemoradiotherapy for thoracic esophageal carcinoma. Int J Radiat Oncol Biol Phys 2009;75:122–8. [4] Ishikura S, Nihei K, Ohtsu A, Boku N, Hironaka S, Mera K, et al. Long-term toxicity after definitive chemoradiotherapy for squamous cell carcinoma of the thoracic esophagus. J Clin Oncol 2003;21:2697–702. [5] Martel MK, Sahijdak WM, ten Haken RK, Kessler ML TA. Fraction size and dose parameters related to the incidence of pericardial effusions. Int J Radiat Oncol Biol Phys 1998;40:155–61. [6] Wei X, Liu HH, Tucker SL, Wang S, Mohan R, Cox JD, et al. Risk factors for pericardial effusion in inoperable esophageal cancer patients treated with definitive chemoradiation therapy. Int J Radiat Oncol Biol Phys 2008;70:707–14. [7] Shirai K, Tamaki Y, Kitamoto Y, Murata K, Satoh Y, Higuchi K, et al. Dose-volume histogram parameters and clinical factors associated with pleural effusion after chemoradiotherapy in esophageal cancer patients. Int J Radiat Oncol Biol Phys 2011;80:1002–7. [8] Mukherjee S, Aston D, Minett M, Brewster a. ., Crosby TD. The Significance of Cardiac Doses Received During Chemoradiation of Oesophageal and Gastro–oesophageal Junctional Cancers. Clin Oncol 2003;15:115–20. [9] Tripp P, Malhotra HK, Javle M, Shaukat a, Russo R, De Boer S, et al. Cardiac function after chemoradiation for esophageal cancer: comparison of heart dose-volume histogram parameters to multiple gated acquisition scan changes. Dis Esophagus 2005;18:400–5. [10] Gayed I, Gohar S, Liao Z, McAleer M, Bassett R, Yusuf SW. The clinical implications of myocardial perfusion abnormalities in patients with esophageal or lung cancer after chemoradiation therapy. Int J Cardiovasc Imaging 2009;25:487–95. [11] Gayed IW, Liu HH, Yusuf SW, Komaki R, Wei X, Wang X, et al. The prevalence of myocardial ischemia after concurrent chemoradiation therapy as detected by gated myocardial perfusion imaging in patients with esophageal cancer. J Nucl Med 2006;47:1756–62. [12] Konski A, Li T, Christensen M, Cheng JD, Yu JQ, Crawford K, et al. Symptomatic cardiac toxicity is predicted by dosimetric and patient factors rather than changes in 18F-FDG PET determination of myocardial activity after chemoradiotherapy for esophageal cancer. Radiother Oncol 2012;104:72–7. [13] Hatakenaka M, Yonezawa M, Nonoshita T, Nakamura K, Yabuuchi H, Shioyama Y, et al. Acute cardiac impairment associated with concurrent chemoradiotherapy for esophageal cancer: magnetic resonance evaluation. Int J Radiat Oncol Biol Phys 2012;83:e67–73. [14] Jingu K, Kaneta T, Nemoto K, Ichinose A, Oikawa M, Takai Y, et al. The utility of 18F-fluorodeoxyglucose positron emission tomography for early diagnosis of radiation-induced myocardial damage. Int J Radiat Oncol Biol Phys 2006;66:845–51. [15] Sjoquist KM, Burmeister BH, Smithers BM, Zalcberg JR, Simes RJ, Barbour A, et al. Survival after neoadjuvant chemotherapy or chemoradiotherapy for resectable oesophageal carcinoma: an updated meta-analysis. Lancet Oncol 2011;12:681–92. [16] Gebski V, Burmeister B, Smithers BM, Foo K, Zalcberg J, Simes J. Survival benefits from neoadjuvant chemoradiotherapy or chemotherapy in oesophageal carcinoma: a meta-analysis. Lancet Oncol 2007;8:226–34. [17] Urschel JD, Vasan H. A meta-analysis of randomized controlled trials that compared neoadjuvant chemoradiation and surgery to surgery alone for resectable esophageal cancer. Am J Surg 2003;185:538–43.

31 Review on cardiac toxicity [18] Van Hagen P, Hulshof MCCM, van Lanschot JJB, Steyerberg EW, van Berge Henegouwen MI, Wijnhoven BPL, et al. Preoperative chemoradiotherapy for esophageal or junctional cancer. N Engl J Med 2012;366:2074–84. [19] Mariette C, Seitz JF, Maillard E, Mornex F, Thomas PA, Raoul J, Boige V, Pezet D, Genet C BL and FF de CD. Surgery alone versus chemoradiotherapy followed by surgery for localised esophageal cancer : analysis od a randomised controlled phase III trial FFCD 9901. J Clin Oncol 2010;28:abstract 4005. [20]Bosch DJ, Muijs CT, Mul VEM, Beukema JC, Hospers G a P, Burgerhof JGM, et al. Impact of Neoadjuvant Chemoradiotherapy on Postoperative Course after Curative-intent Transthoracic Esophagectomy in Esophageal Cancer Patients. Ann Surg Oncol 2013;21. [21] “Port Meta-analysis Trialists group.”Postoperative radiotherapy in non-small-cell lung cancer: systematic review and meta-analysis of individual patient data from nine randomised controlled trials. PORT Meta-analysis Trialists Group. Lancet 1998;352:257–63. [22] Stewart F a. Mechanisms and dose-response relationships for radiation-induced cardiovascular disease. Ann ICRP 2011;41:72–9. [23]Enzinger PC, Mayer RJ. Esophageal cancer. N Engl J Med 2003;349:2241–52. [24]Polk A, Vaage-Nilsen M, Vistisen K, Nielsen DL. Cardiotoxicity in cancer patients treated with 5-fluorouracil or capecitabine: A systematic review of incidence, manifestations and predisposing factors. Cancer Treat Rev 2013;39:974–84. [25]Nuver J, Smit a J, Sleijfer DT, van Gessel a I, van Roon a M, van der Meer J, et al. Left ventricular and cardiac autonomic function in survivors of testicular cancer. Eur J Clin Invest 2005;35:99– 103. [26]Altena R, de Haas EC, Nuver J, Brouwer C a J, van den Berg MP, Smit a J, et al. Evaluation of sub-acute changes in cardiac function after cisplatin-based combination chemotherapy for testicular cancer. Br J Cancer 2009;100:1861–6. [27]Komamura K. Similarities and Differences between the Pathogenesis and Pathophysiology of Diastolic and Systolic Heart Failure. Cardiol Res Pract 2013;2013:824135. [28]Barjaktarovic Z, Shyla A, Azimzadeh O, Schulz S, Haagen J, Dörr W, et al. Ionising radiation induces persistent alterations in the cardiac mitochondrial function of C57BL/6 mice 40 weeks after local heart exposure. Radiother Oncol 2013;106:404–10. [29]Gabriels K, Hoving S, Seemann I, Visser NL, Gijbels MJ, Pol JF, et al. Local heart irradiation of ApoE(-/-) mice induces microvascular and endocardial damage and accelerates coronary atherosclerosis. Radiother Oncol 2012;105:358–64. [30]Seemann I, Gabriels K, Visser NL, Hoving S, te Poele J a, Pol JF, et al. Irradiation induced modest changes in murine cardiac function despite progressive structural damage to the myocardium and microvasculature. Radiother Oncol 2012;103:143–50. [31] Aleman BMP, van den Belt-Dusebout AW, Klokman WJ, Van’t Veer MB, Bartelink H, van Leeuwen FE. Long-term cause-specific mortality of patients treated for Hodgkin’s disease. J Clin Oncol 2003;21:3431–9. [32]McGale P, Darby SC, Hall P, Adolfsson J, Bengtsson N-O, Bennet AM, et al. Incidence of heart disease in 35,000 women treated with radiotherapy for breast cancer in Denmark and Sweden. Radiother Oncol 2011;100:167–75. [33]Lorenzen EL, Taylor CW, Maraldo M, Nielsen MH, Offersen B V, Andersen MR, et al. Interobserver variation in delineation of the heart and left anterior descending coronary artery in radiotherapy for breast cancer: a multi-centre study from Denmark and the UK. Radiother Oncol 2013;108:254–8. [34]Langendijk J a, Lambin P, De Ruysscher D, Widder J, Bos M, Verheij M. Selection of patients for radiotherapy with protons aiming at reduction of side effects: the model-based approach. Radiother Oncol 2013;107:267–73. 2

Made with FlippingBook

RkJQdWJsaXNoZXIy MTk4NDMw