- Email: [email protected]
Subclinical Cardiotoxicity Detected by Strain Rate Imaging up to 14 months After Breast Radiation Therapy
International Journal of
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation: Breast Cancer
Subclinical Cardiotoxicity Detected by Strain Rate Imaging up to 14 months After Breast Radiation Therapy Katrien Erven, MD,*,y Anca Florian, MD,z,x Pieter Slagmolen, PhD,jj,{ Caroline Sweldens, MD,* Ruxandra Jurcut, MD, PhD,x Hans Wildiers, MD, PhD,** Jens-Uwe Voigt, MD, PhD,z and Caroline Weltens, MD, PhD* *Department of Radiotherapy, zDepartment of Cardiology, jjMedical Image Computing (ESAT/PSI), and **Department of Medical Oncology, University Hospital Gasthuisberg, Leuven, Belgium; yIridium Cancer Network, Antwerp, Belgium; x Institute of Emergency for Cardiovascular Diseases, UMF “Carol Davila,” Bucharest, Romania; and {IBBT-KU Leuven Future Health Department, Leuven, Belgium Received Jun 11, 2012, and in revised form Sep 14, 2012. Accepted for publication Sep 18, 2012
Summary Cardiac evaluation with strain rate imaging was performed in 75 breast cancer patients before radiation therapy (RT) and up to 14 months after RT. A significant post-RT reduction in cardiac function was observed for left-sided but not for right-sided patients. Changes were more pronounced in the left ventricular wall receiving the highest dose (anterior wall) compared with the left ventricular wall receiving the lowest dose (inferior wall).
Purpose: Strain rate imaging (SRI) is a new echocardiographic modality that enables accurate measurement of regional myocardial function. We investigated the role of SRI and troponin I (TnI) in the detection of subclinical radiation therapy (RT)-induced cardiotoxicity in breast cancer patients. Methods and Materials: This study prospectively included 75 women (51 left-sided and 24 right-sided) receiving adjuvant RT to the breast/chest wall and regional lymph nodes. Sequential echocardiographs with SRI were obtained before RT, immediately after RT, and 8 and 14 months after RT. TnI levels were measured on the first and last day of RT. Results: Mean heart and left ventricle (LV) doses were both 9 4 Gy for the left-sided patients and 4 4 Gy and 1 0.4 Gy, respectively, for the right-sided patients. A decrease in strain was observed at all post-RT time points for left-sided patients (17.5% 1.9% immediately after RT, 16.6% 1.4% at 8 months, and 17.7% 1.9% at 14 months vs 19.4% 2.4% before RT, P<.01) but not for right-sided patients. When we considered left-sided patients only, the highest mean dose was given to the anterior left ventricular (LV) wall (25 14 Gy) and the lowest to the inferior LV wall (3 3 Gy). Strain of the anterior wall was reduced after RT (16.6% 2.3% immediately after RT, 16% 2.6% at 8 months, and 16.8% 3% at 14 months vs 19% 3.5% before RT, P<.05), whereas strain of the inferior wall showed no significant change. No changes were observed with conventional echocardiography. Furthermore, mean TnI levels for the leftsided patients were significantly elevated after RT compared with before RT, whereas TnI levels of the right-sided patients remained unaffected. Conclusions: In contrast to conventional echocardiography, SRI detected a regional, subclinical decline in cardiac function up to 14 months after breast RT. It remains to be determined whether these changes are related to clinical outcome. In the meantime, we encourage the use of radiation techniques that minimize the exposure of the anterior LV wall in left-sided patients. Ó 2013 Elsevier Inc.
Reprint requests to: Katrien Erven, MD, Department of Radiotherapy, University Hospital Gasthuisberg, Herestraat 49, 3000 Leuven, Belgium. Tel: (þ32) 16347600; Fax: (þ32) 16347623; E-mail: katrien.erven@ uzleuven.be Int J Radiation Oncol Biol Phys, Vol. 85, No. 5, pp. 1172e1178, 2013 0360-3016/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2012.09.022
Katrien Erven, MD and Anca Florian, MD are coauthors on this paper. Conflict of interest: none.
Volume 85 Number 5 2013
Introduction Adjuvant radiation therapy (RT) for breast cancer reduces the risk of local recurrence and can reduce the risk of death from breast cancer (1). However, breast RT is associated with long-term cardiac toxicity (2). Overall, the relative risk of clinically significant cardiac events after RT is within the range of 1.2 to 3.5 (3). In most studies, the endpoint for the evaluation of cardiotoxicity is cardiac morbidity or death. Clinical detection of cardiotoxicity is only relevant 10 to 15 years after RT. Consequently, most data are derived from retrospective studies of patients treated with outdated techniques, exposing large volumes of the heart to relatively high doses. Modern RT techniques enable improved dose distributions, but studies using these techniques lack longterm follow-up, so clinical endpoints such as cardiac morbidity and death might not be adequate. Early surrogate endpoints, predictive for the development of later cardiac events, would be useful. Left ventricular ejection fraction (LVEF) measurement is mostly used to monitor cardiac function. However, LVEF can underestimate actual cardiac damage because of the compensatory reserve of the myocardium that enables adequate ventricular outcome even in the presence of dysfunctional myocytes (4). Moreover, because only part of the left ventricle (LV) is included in the RT fields, mostly regional dysfunction can be expected, which blurs the sensitivity of this global parameter to detect early functional changes after RT. Developments in echocardiographic tissue Doppler imaging, such as strain rate imaging (SRI), have allowed more accurate assessment of regional myocardial function. Several studies have shown that SRI can detect reduced regional myocardial function before any measurable changes in conventional echocardiography parameters occur (5-9). Our group performed a pilot study on the use of SRI for the evaluation of early RT-induced cardiac changes in breast cancer patients. We could show a dose-related regional decrease in myocardial function after RT. However, this small study (30 patients) was limited by its short follow-up of only 2 months (9). Furthermore, cardiac troponins have emerged as very sensitive and specific markers of minor myocardial damage. It has been shown that elevations of serum troponin levels in asymptomatic patients after administration of anthracyclines can predict subsequent deterioration of myocardial function (4). Data on the role of troponins in the detection of RT-induced cardiotoxicity are scarce (10). In this study we investigate the role of SRI in the detection of subclinical RT-induced cardiotoxicity in a larger group of breast cancer patients, having 14 months of follow-up. This will allow us to verify whether the previously observed reductions in cardiac function 2 months after RT are reversible. Furthermore, we assess whether RT-induced elevations of serum troponin levels can be detected.
Methods and Materials Patient population From April 2008 until May 2010, 75 breast cancer patients (51 left-sided and 24 right-sided) (containing none of the pilot study patients) were consecutively enrolled in this prospective study at the University Hospital Leuven. Inclusion criteria were patients
Subclinical radiation-induced cardiotoxicity 1173 with histologically proven early breast carcinoma, requiring adjuvant RT and fulfilling the following criteria: prior surgery (lumpectomy or mastectomy with or without axillary dissection) and prior adjuvant anthracycline- and taxane-based chemotherapy. Exclusion criteria were prior RT with inclusion of the heart in the RT fields; history of serious cardiac illness, including congestive heart failure or cardiomyopathy; and poor echogenicity. The study protocol was approved by the hospital’s ethics committee, and all patients provided written informed consent.
RT technique A computed tomography scan was performed in all patients for RT planning (Somatom Sensation Open; Siemens Medical Solutions, Erlangen, Germany). Breast RT was performed with tangential photon beams. For chest wall RT, we used an anterior electron field. In patients with medially or centrally located tumors or axillary lymph node (LN) invasion, the internal mammary (IM) and medial supraclavicular (MS) LNs were also irradiated. If the IM-MS LNs were treated in combination with the chest wall, an anterior mixed photon and electron beam was used. In case of combination with breast RT, an oblique photon beam for the IM LNs and an anterior photon beam for the MS LNs were used (11). The axilla was only included in the target volume if the axillary LNs were massively involved. Three-dimensional dose calculation was performed with the Eclipse planning system (Varian Medical Systems, Palo Alto, CA). Dose prescription was 50 Gy in 25 fractions. In case of breastconserving surgery, an additional dose was delivered to the tumoral cavity by electrons (16 Gy in 8 fractions) or brachytherapy (8.5 Gy in a single fraction, high dose rate). Because its contribution to the heart dose is neglectable, the boost dose was not considered in the analysis.
Cardiac evaluation Standard echocardiography was performed before RT, immediately after 50 Gy of RT, and at 8 and 14 months after RT, with a Vivid 7 ultrasound scanner (GE Vingmed, Horten, Norway) equipped with a 2.5-MHz broadband transducer. In addition to each conventional echocardiographic examination, tissue Doppler data were acquired from an apical 2-, 3-, and 4-chamber view. Care was taken to maximize frame rate and to align cardiac walls with the ultrasound beam direction. Echocardiographic measurements were performed offline on an EchoPAC workstation (GE Vingmed). To obtain regional information on cardiac function, the LV was divided into 6 walls (anteroseptal, anterior, anterolateral, inferolateral, inferior, and inferoseptal), which were further subdivided into apical, mid, and basal segments, resulting in 18 LV segments per patient. Segmental longitudinal peak systolic strain and strain rate (SR) were extracted from the Doppler myocardial imaging data and analyzed by use of dedicated research software (SPEQLE; University of Leuven, Leuven, Belgium) for all LV segments. It should be noted that strain and SR values are negative, and in the following text, the changes in strain and SR refer to changes in absolute values. Blood samples for cardiac troponin I (TnI) levels were obtained on the first and last day of RT. TnI levels were measured by immunoassay (Dimension RxL; Siemens Medical Solutions). The quantification limit of the assay is 0.13 mg/L.
1174 Erven et al.
International Journal of Radiation Oncology Biology Physics
Radiation dose evaluation and dose-function relations
Results Study population
Dose-volume histogram data were analyzed for the volume of the heart and LV receiving 30 Gy, as well as for the mean heart and LV dose. Normal tissue complication probabilities (NTCPs) were calculated, applying the relative seriality model. The endpoint for NTCP calculation for the heart was excess of cardiac mortality risk (12). To estimate the regional myocardial RT dose, the 3-dimensional RT planning data were imported into a reconstruction software package (MeVisLab 2.1; MeVis Medical Solutions, Bremen, Germany). Within this software, the LV was oriented by manually indicating 3 pointsd1 at the apex, 1 at the middle of the base, and 1 at the middle of the septumdand then automatically segmented into 6 LV walls with 3 segments each, as used in echocardiography (Fig. 1). Mean and maximal segmental RT doses were obtained and compared with time course and spatial distribution of segmental myocardial strain and SR.
Statistics Data are presented as percentages or mean standard deviation. Analysis of differences in means between parameters and time points was performed by use of t tests or analysis of variance with Bonferroni correction for multiple comparisons. Differences in proportions were analyzed with the c2 test with Yates correction. Conventional echocardiographic and segmental SRI parameters were grouped per patient before they were compared between baseline and follow-up visits. Patient- and treatment-related influence on the reduction of regional myocardial function was analyzed by univariate statistical methods. For continuous variables, the Spearman coefficient of correlation was computed, and for categorical variables, the Mann-Whitney test was used. All tests were 2-sided, and a P value below .05 was considered statistically significant. A relatively low intraobserver and interobserver variability (11.8%-14.4%) for strain and SR measurements was reported earlier from our laboratory (13).
Baseline patient characteristics are summarized in Table 1. There was no significant difference between the left- and right-sided patient groups with respect to age, cardiac risk factors, RT target volumes, or adjuvant systemic treatment. IM LN RT was performed in 94% of left-sided and 96% of right-sided patients. All patients received anthracycline- and taxane-based chemotherapy before RT. In all HER2-positive (FISHþ) high-risk patients (22% of left-sided and 17% of right-sided patients), adjuvant treatment with trastuzumab was administered during 1 year after RT. Hormonal treatment was prescribed in case of hormone receptor positivity (76% of left-sided and 88% of right-sided patients).
Follow-up Among the 75 patients enrolled, conventional echocardiography was performed in 74 patients at baseline, 69 patients immediately after RT, 70 patients at 8 months after RT, and 63 patients at 14 months after RT. Additional SRI measurements were performed in 62 patients at baseline, 63 patients immediately after RT, 63 patients at 8 months after RT, and 63 patients at 14 months after RT. TnI assessment was performed in 65 patients before RT and in 59 patients after RT. Both pre- and post-RT TnI values were available for 56 patients.
Cardiac dosimetry and NTCP Table 2 presents the cardiac dose-volume histogram and NTCP values of the left- and right-sided patient groups, as well as mean doses to the different LV walls and apical, mid, and basal segments of the left-sided patients.
Conventional echocardiography Echocardiography parameters are presented in Table 3. We found no significant decrease in conventional echocardiography
Fig. 1. Left ventricular (LV) volume, reconstructed from planning computed tomography (CT) with indication of 3 points (mid-base, apex, and mid-septum) defining the plane, which is used to divide the LV into 6 walls. Each wall is further divided into basal, mid, and apical segments. (A) View at the base of the LV showing the subdivision into 6 walls (anteroseptal [AntSept], anterior [Ant], anterolateral [AntLat], inferolateral [InfLat], inferior [Inf], and inferoseptal [InfSept]). (B) View along the basal-apical axis showing the subdivision into basal, mid, and apical segments.
Volume 85 Number 5 2013 Table 1
Subclinical radiation-induced cardiotoxicity 1175 Table 2
Patient characteristics Left-sided Right-sided P (nZ51) (nZ24) value*
Age (mean SD) (y) Cardiac risk factor [n (%)] Smoker Diabetes mellitus AHT Hypercholesterolemia BMI (mean SD) (kg/m2) RT breast/chest wall [n (%)] Breast Chest wall RT LNs [n (%)] No MS LNs IM-MS LNs IM-MS-AX LNs Chemotherapy [n (%)] FEC-D AC-P CC-P Trastuzumab [n (%)] No Yes Hormonotherapy [n (%)] No Tamoxifen Aromatase inhibitor
54 8
52 7
.19
15 1 15 8 26
5 1 5 1 25
(21) (4) (21) (4) 3
.61 .58 .61 .29 .08
24 (47) 27 (53)
12 (50) 12 (50)
.81
1 2 44 4
0 1 21 2
(0) (4) (88) (8)
.49 .96 .88 .85
47 (92) 2 (4) 2 (4)
21 (88) 1 (4) 2 (8)
.82 .96 .81
40 (78) 11 (22)
20 (83) 4 (17)
.85
12 (24) 19 (37) 20 (39)
3 (12) 10 (42) 11 (46)
.42 .91 .77
(29) (2) (29) (16) 4
(2) (4) (86) (8)
Abbreviations: AC-P Z doxorubicin-cyclophosphamide-paclitaxel; AHT Z arterial hypertension; AX Z axillary; BMI Z body mass index; CC-P Z pegylated liposomal doxorubicin-cyclophosphamide-paclitaxel; FEC-D Z fluorouracil-epirubicin-cyclophosphamide-docetaxel; IM Z internal mammary; LN Z lymph node; MS Z median subclavicular; RT Z radiation therapy. * Student t test for continuous variables and c2 test with Yates correction for proportion data.
parameters for systolic (LVEF, MAPSE) or diastolic (E/A ratio, EDT) function in either left- or right-sided patients at any post-RT time point.
Strain rate imaging Strain and SR assessments were feasible for 988 of 1116 LV segments (89%) at baseline, 1021 of 1134 (90%) immediately after RT, 1003 of 1134 (88%) at 8 months’ follow-up, and 1007 of 1134 (89%) at 14 months’ follow-up. A significant decrease in strain and SR was observed at the different time points after RT for left-sided patients but not for right-sided patients (Table 3, Fig. 2). The largest decrease in strain was noted 8 months after RT. At this time point, the decrease in strain was significantly different between left- and right-sided patients. In left-sided patients, the highest mean dose was given to the anterior wall segments of the LV (25 14 Gy) and the lowest to the inferior wall segments of the LV (3 3 Gy). Mean segmental strain and SR values for the anterior and inferior wall
Heart dose-volume and NTCP data Parameter
Heart Mean dose (Gy) V30 (%) NTCP (%) Left ventricle Mean dose (Gy) V30 (%) Mean dose by location (Gy) Anteroseptal wall Anterior wall Anterolateral wall Inferolateral wall Inferior wall Inferoseptal wall Apical segments Mid segments Basal segments
Left-sided (nZ51)
Right-sided (nZ24)
94 75 1.8 1.7
44 34 0.8 1.4
94 87 14 25 15 4 3 5 10 11 12
9 14 11 3 3 4 12 12 11
1 0.4 00 -
Abbreviations: NTCP Z normal tissue complication probability; V30 (%) Z percent of total volume receiving 30 Gy. Data are given as mean standard deviation.
LV segments of the left-sided patients are shown in Table 3 and Figure 3. Compared with baseline, strain in the anterior segments was significantly decreased at all post-RT time points. In addition, a significant reduction in SR was observed 8 and 14 months after RT. In contrast to the anterior segments, deformation parameters for inferior segments showed no significant changes after RT. Post-RT and 14-month strain values were significantly different between anterior and inferior segments. Univariate analysis verified the effect of different patient- and treatment-related factors on change in strain 14 months after RT. No significant correlations were found between the reduction in strain and the use of trastuzumab, cardiac risk factors, patient age, or cardiac dosimetry. Only a nearly significant, weak correlation could be found with the maximal LV dose (rZ0.33, PZ.06).
TnI assessment Mean TnI levels for the left-sided patients were significantly elevated after RT compared with before RT (0.021 0.01 mg/L before RT and 0.027 0.02 mg/L after RT, PZ.038). For the right-sided patients, no difference was seen between pre- and postRT values (0.020 0.02 mg/L before RT and 0.021 0.01 mg/L after RT, PZ.8). In all patients the pre- and post-RT TnI levels were below the clinically used cutoff value of 0.13 mg/L.
Discussion Main findings In this study SRI was used to measure RT-induced changes in myocardial function in patients receiving breast or chest wall RT. In nearly all patients, IM LN RT was performed, and all patients were pretreated with anthracycline- and taxane-based
International Journal of Radiation Oncology Biology Physics
1176 Erven et al. Table 3
Evolution of echocardiographic parameters over follow-up
LVEF (%) Left Right MAPSE (mm) Left Right E/A ratio Left Right EDT (ms) Left Right S (%) Left Right SR (1/s) Left Right S (%) Anterior wallx Inferior wallx SR (1/s) Anterior wallx Inferior wallx
Before RT
After RT
8 mo
14 mo
61.4 3.9 62.6 3.6
60 3.7 61.8 4.2
60 2.9 61.9 4
60.3 3.4 61.2 3.5
12.3 1.8 12.9 1.7
12.6 1.5 13.1 1.8
12.6 1.3 13.2 1.4
12.3 1.7 13.2 1.8
1.1 0.3 1.1 0.3
1 0.2 1 0.2
1 0.3 1 0.2
1 0.3 1 0.2
192 25 180 29
195 35 182 29
194 30 184 29
198 29 188 16
19.4 2.4 18.9 1.8
17.5 1.9* 18.4 1.5
16.6 1.4y 17.7 1.8
17.7 1.9* 18.2 2.5
1.4 0.26 1.36 0.12
1.22 0.15* 1.24 0.13
1.23 0.17* 1.28 0.1
1.25 0.17* 1.29 0.17
19 3.5 18.8 2.7
16.6 2.3z 18.9 3.1
16 2.6* 17.1 2.8
16.8 3z 18.4 3
1.49 0.41 1.26 0.3
1.29 0.27 1.21 0.24
1.19 0.25* 1.11 0.21
1.27 0.26z 1.19 0.23
Abbreviations: LVEF Z left ventricular ejection fraction; MAPSE Z mitral annular plane of systolic excursion; EDT Z E-wave deceleration time; RT Z radiation therapy; S Z strain; SR Z strain rate. Data are given as mean standard deviation. * P<.01 for comparison with baseline (repeated-measures analysis). y P<.001 for comparison with baseline (repeated-measures analysis). z P<.05 for comparison with baseline (repeated-measures analysis). x Only left-sided patients are considered.
chemotherapy. A significant decrease in longitudinal strain and SR was observed immediately after RT and at 8 and 14 months after RT for left-sided breast cancer patients but not for right-sided patients. Changes were more pronounced in the LV wall receiving the highest RT dose (anterior wall) compared with the LV wall receiving the lowest RT dose (inferior wall).
Our data confirm, in a larger number of patients, our previously reported findings that SRI can identify reductions in LV function immediately after RT that are not detectable by conventional echocardiographic function parameters (9). In addition, this study shows that the cardiac function remains affected up to 14 months after RT.
Fig. 2. Course of mean longitudinal strain (A) and strain rate (B) in left- and right-sided patients at the different time points. P values of significant differences between baseline and follow-up examinations for the left-sided patients are shown above the curves. P values of significant differences between both groups (left- and right-sided) are shown between the curves. RT Z radiation therapy.
Volume 85 Number 5 2013
Subclinical radiation-induced cardiotoxicity 1177
Fig. 3. Course of mean longitudinal strain (A) and strain rate (B) in anterior and inferior left ventricular segments of left-sided patients. P values of significant differences between baseline and follow-up examinations for the anterior segments are shown above or beneath the curves. P values of significant differences between both groups (anterior and inferior segments) are shown between the curves. RT Z radiation therapy.
Use of SRI in detection of treatment-induced cardiotoxicity To our knowledge, the detection of RT-induced cardiotoxicity by SRI has not yet been studied by other groups. However, limited but promising data are available on the use of SRI in the detection of anthracycline- or trastuzumab-induced cardiotoxicity. All studies conclude that simply measuring LVEF is not sensitive enough for the detection of subclinical cardiotoxicity and encourage the use of SRI (4-9). Two studies, using SRI in the detection of anthracycline- or trastuzumab-induced cardiotoxicity, investigated whether RT had an influence on the observed reductions in cardiac function caused by the systemic treatments (7, 8). Both studies detected no interaction of RT with the deformation results. Unfortunately, neither publication provides details on RT heart doses, treatment side, or target volume (ie, inclusion of LNs), and the findings are assumedly based on a comparison between the group of patients receiving RT and the group receiving no RT, without taking into account the side of RT. As we can show in our study, hardly any dose is received by the LV in right-sided breast or chest wall RT and myocardial function changes clearly differ depending on the side of RT. Therefore we think that a possible interaction of RT cannot be excluded when not only left-sided patients are considered.
Effect of chemotherapy on strain/SR reductions Because all patients in our study were treated with anthracyclinebased chemotherapy before RT, a late chemotherapy effect on myocardial function cannot be ruled out. However, no significant post-RT reductions in strain/SR were seen in right-sided patients or in myocardial segments of left-sided patients receiving a low radiation dose (inferior wall). We are therefore confident that the observed differences in strain and SR are mainly due to RT.
Effect of radiation dose and volume on strain/SR reductions The reported differences in reduction of post-RT strain/SR between left- and right-sided patients and between the anterior
and inferior LV wall in left-sided patients suggest a relevant impact of RT dose and volume. However, with our data, we could not show a statistically significant relation between segmental RT dose and segmental function changes after RT. Consequently, no threshold dose level for increased cardiac toxicity risk could be proposed. We suggest that this counterintuitive finding should be interpreted as resulting from the uncertainty regarding the true RT dose to a small region of the myocardium, which could only be estimated but not measured. This explanation is strengthened by the finding that significant functional differences could be observed when larger myocardial areas (anterior wall vs inferior wall) were considered. Furthermore, apart from myocardial segments, other cardiac substructures, such as the coronary arteries, could also be relevant exposure structures, blurring the relation between segmental dose and function (4). Further study is needed to examine which part of the heart is most radiosensitive and should be used in the search for tolerance doses below which there is no increased cardiac risk. Meanwhile, we think it is worthwhile to reduce the dose to the heart as much as possible using modern radiation techniques such as respiratory gating.
TnI assessment Cardiac biomarkers, particularly troponins, have extensively been studied as a tool to evaluate chemotherapy cardiotoxicity. Elevations of serum cardiac troponin levels in asymptomatic patients after administration of anthracyclines predict deterioration of myocardial function (4). In contrast, very limited data exist on the role of troponins in the early detection of RT-induced cardiotoxicity. Hughes-Davies et al (10) quantified serum troponin T levels in 50 left-sided breast cancer patients before and after RT. They did not find any change in serum troponin T levels after RT. The patient population of this study differs from ours in that IM LNs were not irradiated, the total RT dose was lower, and none of the patients were previously treated with anthracyclines or trastuzumab. In our study, TnI levels were significantly elevated after RT compared with before RT in the left-sided group, whereas in the right-sided group, no change in TnI levels could be noticed. However, in all patients the TnI values remained below the quantification level. Therefore our findings could still be
1178 Erven et al.
International Journal of Radiation Oncology Biology Physics
coincidental, because the assay that was used is possibly not sensitive enough to guarantee the results of the TnI assessment below the quantification level.
Therefore we encourage the use of new techniques, such as respiratory gated RT, enabling reduced exposure of the anterior LV wall.
Limitations Interpatient variations in cardiotoxic risk may be caused by several patient- and treatment-related factors. Larger patient groups are needed to increase the power to detect the influence of other factors on post-RT strain/SR reductions. SRI should also be performed before the start of chemotherapy to be able to define the effect of chemotherapy on the observed strain/SR reductions. Dose-volume characteristics were calculated for 18 LV segments but not for the coronary arteries. Because the coronary arteries have a very small volume and no contrast was given for the planning computed tomography scan, exact contouring of the coronary arteries is problematic and prevents meaningful dosevolume statistics. The relevance of strain/SR reductions as a surrogate parameter for clinical long-term effects of RT on myocardial function needs further clarification. Radiation-induced cardiac disease may become clinically manifest only 10 years or more after treatment, so further investigation with long-term follow-up is needed to examine whether patients with larger reduction in strain/SR are more likely to have cardiac dysfunction. The immunoassay that was used for TnI measurement had a rather high quantification level and was therefore not ideal to use for detection of acute post-RT cardiotoxicity, which is expected to be only minor myocardial damage. To confirm the differences in TnI levels that were found after RT in the left-sided patient group and assess the value of troponins in the early detection of post-RT cardiotoxicity, a high-sensitivity assay should be used.
Conclusions SRI is a sensitive method to detect RT-induced changes in cardiac function that are not detectable by conventional echocardiographic measures. These subclinical reductions in cardiac function persist up to 14 months after RT. Further investigations should define whether SRI can be predictive for late cardiac morbidity. It might be an interesting tool to define patients at risk, who could possibly be saved by the early initiation of preventive measures. A critical dose level for cardiac dysfunction could not be proposed. Post-RT reduction in cardiac function was most prominent in the anterior LV wall of left-sided breast cancer patients.
References 1. Clarke M, Collins R, Darby S, et al. Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: An overview of the randomised trials. Lancet 2005;366:2087-2106. 2. Cuzick J, Stewart H, Rutqvist L, et al. Cause-specific mortality in long-term survivors of breast cancer who participated in trials of radiotherapy. J Clin Oncol 1994;12:447-453. 3. Gagliardi G, Constine LS, Moiseenko V, et al. Radiation dosevolume effects in the heart. Int J Radiat Oncol Biol Phys 2010; 76:S77-S85. 4. Altena R, Perik PJ, van Veldhuisen DJ, de Vries EG, Gietema JA. Cardiovascular toxicity caused by cancer treatment: Strategies for early detection. Lancet Oncol 2009;10:391-399. 5. Ganame J, Claus P, Eyskens B, et al. Acute cardiac functional and morphological changes after Anthracycline infusions in children. Am J Cardiol 2007;99:974-977. 6. Jurcut R, Wildiers H, Ganame J, et al. Strain rate imaging detects early cardiac effects of pegylated liposomal doxorubicin as adjuvant therapy in elderly patients with breast cancer. J Am Soc Echocardiogr 2008;21: 1283-1289. 7. Ho E, Brown A, Barrett P, et al. Subclinical anthracycline- and trastuzumab-induced cardiotoxicity in the long-term follow-up of asymptomatic breast cancer survivors: A speckle tracking echocardiographic study. Heart 2010;96:701-707. 8. Hare JL, Brown JK, Leano R, Jenkins C, Woodward N, Marwick TH. Use of myocardial deformation imaging to detect preclinical myocardial dysfunction before conventional measures in patients undergoing breast cancer treatment with trastuzumab. Am Heart J 2009;158:294-301. 9. Erven K, Jurcut R, Weltens C, et al. Acute radiation effects on cardiac function detected by strain rate imaging in breast cancer patients. Int J Radiat Oncol Biol Phys 2011;79:1444-1451. 10. Hughes-Davies L, Sacks D, Rescigno J, Howard S, Harris J. Serum cardiac troponin T levels during treatment of early-stage breast cancer. J Clin Oncol 1995;13:2582-2584. 11. Erven K, Petillion S, Weltens C, et al. Conformal locoregional breast irradiation with an oblique parasternal photon field technique. Med Dosim 2011;36:28-34. 12. Gagliardi G, Lax I, Soderstrom S, Gyenes G, Rutqvist LE. Prediction of excess risk of long-term cardiac mortality after radiotherapy of stage I breast cancer. Radiother Oncol 1998;46:63-71. 13. Kowalski M, Kukulski T, Jamal F, et al. Can natural strain and strain rate quantify regional myocardial deformation? A study in healthy subjects. Ultrasound Med Biol 2001;27:1087-1097.
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation: Breast Cancer
Subclinical Cardiotoxicity Detected by Strain Rate Imaging up to 14 months After Breast Radiation Therapy Katrien Erven, MD,*,y Anca Florian, MD,z,x Pieter Slagmolen, PhD,jj,{ Caroline Sweldens, MD,* Ruxandra Jurcut, MD, PhD,x Hans Wildiers, MD, PhD,** Jens-Uwe Voigt, MD, PhD,z and Caroline Weltens, MD, PhD* *Department of Radiotherapy, zDepartment of Cardiology, jjMedical Image Computing (ESAT/PSI), and **Department of Medical Oncology, University Hospital Gasthuisberg, Leuven, Belgium; yIridium Cancer Network, Antwerp, Belgium; x Institute of Emergency for Cardiovascular Diseases, UMF “Carol Davila,” Bucharest, Romania; and {IBBT-KU Leuven Future Health Department, Leuven, Belgium Received Jun 11, 2012, and in revised form Sep 14, 2012. Accepted for publication Sep 18, 2012
Summary Cardiac evaluation with strain rate imaging was performed in 75 breast cancer patients before radiation therapy (RT) and up to 14 months after RT. A significant post-RT reduction in cardiac function was observed for left-sided but not for right-sided patients. Changes were more pronounced in the left ventricular wall receiving the highest dose (anterior wall) compared with the left ventricular wall receiving the lowest dose (inferior wall).
Purpose: Strain rate imaging (SRI) is a new echocardiographic modality that enables accurate measurement of regional myocardial function. We investigated the role of SRI and troponin I (TnI) in the detection of subclinical radiation therapy (RT)-induced cardiotoxicity in breast cancer patients. Methods and Materials: This study prospectively included 75 women (51 left-sided and 24 right-sided) receiving adjuvant RT to the breast/chest wall and regional lymph nodes. Sequential echocardiographs with SRI were obtained before RT, immediately after RT, and 8 and 14 months after RT. TnI levels were measured on the first and last day of RT. Results: Mean heart and left ventricle (LV) doses were both 9 4 Gy for the left-sided patients and 4 4 Gy and 1 0.4 Gy, respectively, for the right-sided patients. A decrease in strain was observed at all post-RT time points for left-sided patients (17.5% 1.9% immediately after RT, 16.6% 1.4% at 8 months, and 17.7% 1.9% at 14 months vs 19.4% 2.4% before RT, P<.01) but not for right-sided patients. When we considered left-sided patients only, the highest mean dose was given to the anterior left ventricular (LV) wall (25 14 Gy) and the lowest to the inferior LV wall (3 3 Gy). Strain of the anterior wall was reduced after RT (16.6% 2.3% immediately after RT, 16% 2.6% at 8 months, and 16.8% 3% at 14 months vs 19% 3.5% before RT, P<.05), whereas strain of the inferior wall showed no significant change. No changes were observed with conventional echocardiography. Furthermore, mean TnI levels for the leftsided patients were significantly elevated after RT compared with before RT, whereas TnI levels of the right-sided patients remained unaffected. Conclusions: In contrast to conventional echocardiography, SRI detected a regional, subclinical decline in cardiac function up to 14 months after breast RT. It remains to be determined whether these changes are related to clinical outcome. In the meantime, we encourage the use of radiation techniques that minimize the exposure of the anterior LV wall in left-sided patients. Ó 2013 Elsevier Inc.
Reprint requests to: Katrien Erven, MD, Department of Radiotherapy, University Hospital Gasthuisberg, Herestraat 49, 3000 Leuven, Belgium. Tel: (þ32) 16347600; Fax: (þ32) 16347623; E-mail: katrien.erven@ uzleuven.be Int J Radiation Oncol Biol Phys, Vol. 85, No. 5, pp. 1172e1178, 2013 0360-3016/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2012.09.022
Katrien Erven, MD and Anca Florian, MD are coauthors on this paper. Conflict of interest: none.
Volume 85 Number 5 2013
Introduction Adjuvant radiation therapy (RT) for breast cancer reduces the risk of local recurrence and can reduce the risk of death from breast cancer (1). However, breast RT is associated with long-term cardiac toxicity (2). Overall, the relative risk of clinically significant cardiac events after RT is within the range of 1.2 to 3.5 (3). In most studies, the endpoint for the evaluation of cardiotoxicity is cardiac morbidity or death. Clinical detection of cardiotoxicity is only relevant 10 to 15 years after RT. Consequently, most data are derived from retrospective studies of patients treated with outdated techniques, exposing large volumes of the heart to relatively high doses. Modern RT techniques enable improved dose distributions, but studies using these techniques lack longterm follow-up, so clinical endpoints such as cardiac morbidity and death might not be adequate. Early surrogate endpoints, predictive for the development of later cardiac events, would be useful. Left ventricular ejection fraction (LVEF) measurement is mostly used to monitor cardiac function. However, LVEF can underestimate actual cardiac damage because of the compensatory reserve of the myocardium that enables adequate ventricular outcome even in the presence of dysfunctional myocytes (4). Moreover, because only part of the left ventricle (LV) is included in the RT fields, mostly regional dysfunction can be expected, which blurs the sensitivity of this global parameter to detect early functional changes after RT. Developments in echocardiographic tissue Doppler imaging, such as strain rate imaging (SRI), have allowed more accurate assessment of regional myocardial function. Several studies have shown that SRI can detect reduced regional myocardial function before any measurable changes in conventional echocardiography parameters occur (5-9). Our group performed a pilot study on the use of SRI for the evaluation of early RT-induced cardiac changes in breast cancer patients. We could show a dose-related regional decrease in myocardial function after RT. However, this small study (30 patients) was limited by its short follow-up of only 2 months (9). Furthermore, cardiac troponins have emerged as very sensitive and specific markers of minor myocardial damage. It has been shown that elevations of serum troponin levels in asymptomatic patients after administration of anthracyclines can predict subsequent deterioration of myocardial function (4). Data on the role of troponins in the detection of RT-induced cardiotoxicity are scarce (10). In this study we investigate the role of SRI in the detection of subclinical RT-induced cardiotoxicity in a larger group of breast cancer patients, having 14 months of follow-up. This will allow us to verify whether the previously observed reductions in cardiac function 2 months after RT are reversible. Furthermore, we assess whether RT-induced elevations of serum troponin levels can be detected.
Methods and Materials Patient population From April 2008 until May 2010, 75 breast cancer patients (51 left-sided and 24 right-sided) (containing none of the pilot study patients) were consecutively enrolled in this prospective study at the University Hospital Leuven. Inclusion criteria were patients
Subclinical radiation-induced cardiotoxicity 1173 with histologically proven early breast carcinoma, requiring adjuvant RT and fulfilling the following criteria: prior surgery (lumpectomy or mastectomy with or without axillary dissection) and prior adjuvant anthracycline- and taxane-based chemotherapy. Exclusion criteria were prior RT with inclusion of the heart in the RT fields; history of serious cardiac illness, including congestive heart failure or cardiomyopathy; and poor echogenicity. The study protocol was approved by the hospital’s ethics committee, and all patients provided written informed consent.
RT technique A computed tomography scan was performed in all patients for RT planning (Somatom Sensation Open; Siemens Medical Solutions, Erlangen, Germany). Breast RT was performed with tangential photon beams. For chest wall RT, we used an anterior electron field. In patients with medially or centrally located tumors or axillary lymph node (LN) invasion, the internal mammary (IM) and medial supraclavicular (MS) LNs were also irradiated. If the IM-MS LNs were treated in combination with the chest wall, an anterior mixed photon and electron beam was used. In case of combination with breast RT, an oblique photon beam for the IM LNs and an anterior photon beam for the MS LNs were used (11). The axilla was only included in the target volume if the axillary LNs were massively involved. Three-dimensional dose calculation was performed with the Eclipse planning system (Varian Medical Systems, Palo Alto, CA). Dose prescription was 50 Gy in 25 fractions. In case of breastconserving surgery, an additional dose was delivered to the tumoral cavity by electrons (16 Gy in 8 fractions) or brachytherapy (8.5 Gy in a single fraction, high dose rate). Because its contribution to the heart dose is neglectable, the boost dose was not considered in the analysis.
Cardiac evaluation Standard echocardiography was performed before RT, immediately after 50 Gy of RT, and at 8 and 14 months after RT, with a Vivid 7 ultrasound scanner (GE Vingmed, Horten, Norway) equipped with a 2.5-MHz broadband transducer. In addition to each conventional echocardiographic examination, tissue Doppler data were acquired from an apical 2-, 3-, and 4-chamber view. Care was taken to maximize frame rate and to align cardiac walls with the ultrasound beam direction. Echocardiographic measurements were performed offline on an EchoPAC workstation (GE Vingmed). To obtain regional information on cardiac function, the LV was divided into 6 walls (anteroseptal, anterior, anterolateral, inferolateral, inferior, and inferoseptal), which were further subdivided into apical, mid, and basal segments, resulting in 18 LV segments per patient. Segmental longitudinal peak systolic strain and strain rate (SR) were extracted from the Doppler myocardial imaging data and analyzed by use of dedicated research software (SPEQLE; University of Leuven, Leuven, Belgium) for all LV segments. It should be noted that strain and SR values are negative, and in the following text, the changes in strain and SR refer to changes in absolute values. Blood samples for cardiac troponin I (TnI) levels were obtained on the first and last day of RT. TnI levels were measured by immunoassay (Dimension RxL; Siemens Medical Solutions). The quantification limit of the assay is 0.13 mg/L.
1174 Erven et al.
International Journal of Radiation Oncology Biology Physics
Radiation dose evaluation and dose-function relations
Results Study population
Dose-volume histogram data were analyzed for the volume of the heart and LV receiving 30 Gy, as well as for the mean heart and LV dose. Normal tissue complication probabilities (NTCPs) were calculated, applying the relative seriality model. The endpoint for NTCP calculation for the heart was excess of cardiac mortality risk (12). To estimate the regional myocardial RT dose, the 3-dimensional RT planning data were imported into a reconstruction software package (MeVisLab 2.1; MeVis Medical Solutions, Bremen, Germany). Within this software, the LV was oriented by manually indicating 3 pointsd1 at the apex, 1 at the middle of the base, and 1 at the middle of the septumdand then automatically segmented into 6 LV walls with 3 segments each, as used in echocardiography (Fig. 1). Mean and maximal segmental RT doses were obtained and compared with time course and spatial distribution of segmental myocardial strain and SR.
Statistics Data are presented as percentages or mean standard deviation. Analysis of differences in means between parameters and time points was performed by use of t tests or analysis of variance with Bonferroni correction for multiple comparisons. Differences in proportions were analyzed with the c2 test with Yates correction. Conventional echocardiographic and segmental SRI parameters were grouped per patient before they were compared between baseline and follow-up visits. Patient- and treatment-related influence on the reduction of regional myocardial function was analyzed by univariate statistical methods. For continuous variables, the Spearman coefficient of correlation was computed, and for categorical variables, the Mann-Whitney test was used. All tests were 2-sided, and a P value below .05 was considered statistically significant. A relatively low intraobserver and interobserver variability (11.8%-14.4%) for strain and SR measurements was reported earlier from our laboratory (13).
Baseline patient characteristics are summarized in Table 1. There was no significant difference between the left- and right-sided patient groups with respect to age, cardiac risk factors, RT target volumes, or adjuvant systemic treatment. IM LN RT was performed in 94% of left-sided and 96% of right-sided patients. All patients received anthracycline- and taxane-based chemotherapy before RT. In all HER2-positive (FISHþ) high-risk patients (22% of left-sided and 17% of right-sided patients), adjuvant treatment with trastuzumab was administered during 1 year after RT. Hormonal treatment was prescribed in case of hormone receptor positivity (76% of left-sided and 88% of right-sided patients).
Follow-up Among the 75 patients enrolled, conventional echocardiography was performed in 74 patients at baseline, 69 patients immediately after RT, 70 patients at 8 months after RT, and 63 patients at 14 months after RT. Additional SRI measurements were performed in 62 patients at baseline, 63 patients immediately after RT, 63 patients at 8 months after RT, and 63 patients at 14 months after RT. TnI assessment was performed in 65 patients before RT and in 59 patients after RT. Both pre- and post-RT TnI values were available for 56 patients.
Cardiac dosimetry and NTCP Table 2 presents the cardiac dose-volume histogram and NTCP values of the left- and right-sided patient groups, as well as mean doses to the different LV walls and apical, mid, and basal segments of the left-sided patients.
Conventional echocardiography Echocardiography parameters are presented in Table 3. We found no significant decrease in conventional echocardiography
Fig. 1. Left ventricular (LV) volume, reconstructed from planning computed tomography (CT) with indication of 3 points (mid-base, apex, and mid-septum) defining the plane, which is used to divide the LV into 6 walls. Each wall is further divided into basal, mid, and apical segments. (A) View at the base of the LV showing the subdivision into 6 walls (anteroseptal [AntSept], anterior [Ant], anterolateral [AntLat], inferolateral [InfLat], inferior [Inf], and inferoseptal [InfSept]). (B) View along the basal-apical axis showing the subdivision into basal, mid, and apical segments.
Volume 85 Number 5 2013 Table 1
Subclinical radiation-induced cardiotoxicity 1175 Table 2
Patient characteristics Left-sided Right-sided P (nZ51) (nZ24) value*
Age (mean SD) (y) Cardiac risk factor [n (%)] Smoker Diabetes mellitus AHT Hypercholesterolemia BMI (mean SD) (kg/m2) RT breast/chest wall [n (%)] Breast Chest wall RT LNs [n (%)] No MS LNs IM-MS LNs IM-MS-AX LNs Chemotherapy [n (%)] FEC-D AC-P CC-P Trastuzumab [n (%)] No Yes Hormonotherapy [n (%)] No Tamoxifen Aromatase inhibitor
54 8
52 7
.19
15 1 15 8 26
5 1 5 1 25
(21) (4) (21) (4) 3
.61 .58 .61 .29 .08
24 (47) 27 (53)
12 (50) 12 (50)
.81
1 2 44 4
0 1 21 2
(0) (4) (88) (8)
.49 .96 .88 .85
47 (92) 2 (4) 2 (4)
21 (88) 1 (4) 2 (8)
.82 .96 .81
40 (78) 11 (22)
20 (83) 4 (17)
.85
12 (24) 19 (37) 20 (39)
3 (12) 10 (42) 11 (46)
.42 .91 .77
(29) (2) (29) (16) 4
(2) (4) (86) (8)
Abbreviations: AC-P Z doxorubicin-cyclophosphamide-paclitaxel; AHT Z arterial hypertension; AX Z axillary; BMI Z body mass index; CC-P Z pegylated liposomal doxorubicin-cyclophosphamide-paclitaxel; FEC-D Z fluorouracil-epirubicin-cyclophosphamide-docetaxel; IM Z internal mammary; LN Z lymph node; MS Z median subclavicular; RT Z radiation therapy. * Student t test for continuous variables and c2 test with Yates correction for proportion data.
parameters for systolic (LVEF, MAPSE) or diastolic (E/A ratio, EDT) function in either left- or right-sided patients at any post-RT time point.
Strain rate imaging Strain and SR assessments were feasible for 988 of 1116 LV segments (89%) at baseline, 1021 of 1134 (90%) immediately after RT, 1003 of 1134 (88%) at 8 months’ follow-up, and 1007 of 1134 (89%) at 14 months’ follow-up. A significant decrease in strain and SR was observed at the different time points after RT for left-sided patients but not for right-sided patients (Table 3, Fig. 2). The largest decrease in strain was noted 8 months after RT. At this time point, the decrease in strain was significantly different between left- and right-sided patients. In left-sided patients, the highest mean dose was given to the anterior wall segments of the LV (25 14 Gy) and the lowest to the inferior wall segments of the LV (3 3 Gy). Mean segmental strain and SR values for the anterior and inferior wall
Heart dose-volume and NTCP data Parameter
Heart Mean dose (Gy) V30 (%) NTCP (%) Left ventricle Mean dose (Gy) V30 (%) Mean dose by location (Gy) Anteroseptal wall Anterior wall Anterolateral wall Inferolateral wall Inferior wall Inferoseptal wall Apical segments Mid segments Basal segments
Left-sided (nZ51)
Right-sided (nZ24)
94 75 1.8 1.7
44 34 0.8 1.4
94 87 14 25 15 4 3 5 10 11 12
9 14 11 3 3 4 12 12 11
1 0.4 00 -
Abbreviations: NTCP Z normal tissue complication probability; V30 (%) Z percent of total volume receiving 30 Gy. Data are given as mean standard deviation.
LV segments of the left-sided patients are shown in Table 3 and Figure 3. Compared with baseline, strain in the anterior segments was significantly decreased at all post-RT time points. In addition, a significant reduction in SR was observed 8 and 14 months after RT. In contrast to the anterior segments, deformation parameters for inferior segments showed no significant changes after RT. Post-RT and 14-month strain values were significantly different between anterior and inferior segments. Univariate analysis verified the effect of different patient- and treatment-related factors on change in strain 14 months after RT. No significant correlations were found between the reduction in strain and the use of trastuzumab, cardiac risk factors, patient age, or cardiac dosimetry. Only a nearly significant, weak correlation could be found with the maximal LV dose (rZ0.33, PZ.06).
TnI assessment Mean TnI levels for the left-sided patients were significantly elevated after RT compared with before RT (0.021 0.01 mg/L before RT and 0.027 0.02 mg/L after RT, PZ.038). For the right-sided patients, no difference was seen between pre- and postRT values (0.020 0.02 mg/L before RT and 0.021 0.01 mg/L after RT, PZ.8). In all patients the pre- and post-RT TnI levels were below the clinically used cutoff value of 0.13 mg/L.
Discussion Main findings In this study SRI was used to measure RT-induced changes in myocardial function in patients receiving breast or chest wall RT. In nearly all patients, IM LN RT was performed, and all patients were pretreated with anthracycline- and taxane-based
International Journal of Radiation Oncology Biology Physics
1176 Erven et al. Table 3
Evolution of echocardiographic parameters over follow-up
LVEF (%) Left Right MAPSE (mm) Left Right E/A ratio Left Right EDT (ms) Left Right S (%) Left Right SR (1/s) Left Right S (%) Anterior wallx Inferior wallx SR (1/s) Anterior wallx Inferior wallx
Before RT
After RT
8 mo
14 mo
61.4 3.9 62.6 3.6
60 3.7 61.8 4.2
60 2.9 61.9 4
60.3 3.4 61.2 3.5
12.3 1.8 12.9 1.7
12.6 1.5 13.1 1.8
12.6 1.3 13.2 1.4
12.3 1.7 13.2 1.8
1.1 0.3 1.1 0.3
1 0.2 1 0.2
1 0.3 1 0.2
1 0.3 1 0.2
192 25 180 29
195 35 182 29
194 30 184 29
198 29 188 16
19.4 2.4 18.9 1.8
17.5 1.9* 18.4 1.5
16.6 1.4y 17.7 1.8
17.7 1.9* 18.2 2.5
1.4 0.26 1.36 0.12
1.22 0.15* 1.24 0.13
1.23 0.17* 1.28 0.1
1.25 0.17* 1.29 0.17
19 3.5 18.8 2.7
16.6 2.3z 18.9 3.1
16 2.6* 17.1 2.8
16.8 3z 18.4 3
1.49 0.41 1.26 0.3
1.29 0.27 1.21 0.24
1.19 0.25* 1.11 0.21
1.27 0.26z 1.19 0.23
Abbreviations: LVEF Z left ventricular ejection fraction; MAPSE Z mitral annular plane of systolic excursion; EDT Z E-wave deceleration time; RT Z radiation therapy; S Z strain; SR Z strain rate. Data are given as mean standard deviation. * P<.01 for comparison with baseline (repeated-measures analysis). y P<.001 for comparison with baseline (repeated-measures analysis). z P<.05 for comparison with baseline (repeated-measures analysis). x Only left-sided patients are considered.
chemotherapy. A significant decrease in longitudinal strain and SR was observed immediately after RT and at 8 and 14 months after RT for left-sided breast cancer patients but not for right-sided patients. Changes were more pronounced in the LV wall receiving the highest RT dose (anterior wall) compared with the LV wall receiving the lowest RT dose (inferior wall).
Our data confirm, in a larger number of patients, our previously reported findings that SRI can identify reductions in LV function immediately after RT that are not detectable by conventional echocardiographic function parameters (9). In addition, this study shows that the cardiac function remains affected up to 14 months after RT.
Fig. 2. Course of mean longitudinal strain (A) and strain rate (B) in left- and right-sided patients at the different time points. P values of significant differences between baseline and follow-up examinations for the left-sided patients are shown above the curves. P values of significant differences between both groups (left- and right-sided) are shown between the curves. RT Z radiation therapy.
Volume 85 Number 5 2013
Subclinical radiation-induced cardiotoxicity 1177
Fig. 3. Course of mean longitudinal strain (A) and strain rate (B) in anterior and inferior left ventricular segments of left-sided patients. P values of significant differences between baseline and follow-up examinations for the anterior segments are shown above or beneath the curves. P values of significant differences between both groups (anterior and inferior segments) are shown between the curves. RT Z radiation therapy.
Use of SRI in detection of treatment-induced cardiotoxicity To our knowledge, the detection of RT-induced cardiotoxicity by SRI has not yet been studied by other groups. However, limited but promising data are available on the use of SRI in the detection of anthracycline- or trastuzumab-induced cardiotoxicity. All studies conclude that simply measuring LVEF is not sensitive enough for the detection of subclinical cardiotoxicity and encourage the use of SRI (4-9). Two studies, using SRI in the detection of anthracycline- or trastuzumab-induced cardiotoxicity, investigated whether RT had an influence on the observed reductions in cardiac function caused by the systemic treatments (7, 8). Both studies detected no interaction of RT with the deformation results. Unfortunately, neither publication provides details on RT heart doses, treatment side, or target volume (ie, inclusion of LNs), and the findings are assumedly based on a comparison between the group of patients receiving RT and the group receiving no RT, without taking into account the side of RT. As we can show in our study, hardly any dose is received by the LV in right-sided breast or chest wall RT and myocardial function changes clearly differ depending on the side of RT. Therefore we think that a possible interaction of RT cannot be excluded when not only left-sided patients are considered.
Effect of chemotherapy on strain/SR reductions Because all patients in our study were treated with anthracyclinebased chemotherapy before RT, a late chemotherapy effect on myocardial function cannot be ruled out. However, no significant post-RT reductions in strain/SR were seen in right-sided patients or in myocardial segments of left-sided patients receiving a low radiation dose (inferior wall). We are therefore confident that the observed differences in strain and SR are mainly due to RT.
Effect of radiation dose and volume on strain/SR reductions The reported differences in reduction of post-RT strain/SR between left- and right-sided patients and between the anterior
and inferior LV wall in left-sided patients suggest a relevant impact of RT dose and volume. However, with our data, we could not show a statistically significant relation between segmental RT dose and segmental function changes after RT. Consequently, no threshold dose level for increased cardiac toxicity risk could be proposed. We suggest that this counterintuitive finding should be interpreted as resulting from the uncertainty regarding the true RT dose to a small region of the myocardium, which could only be estimated but not measured. This explanation is strengthened by the finding that significant functional differences could be observed when larger myocardial areas (anterior wall vs inferior wall) were considered. Furthermore, apart from myocardial segments, other cardiac substructures, such as the coronary arteries, could also be relevant exposure structures, blurring the relation between segmental dose and function (4). Further study is needed to examine which part of the heart is most radiosensitive and should be used in the search for tolerance doses below which there is no increased cardiac risk. Meanwhile, we think it is worthwhile to reduce the dose to the heart as much as possible using modern radiation techniques such as respiratory gating.
TnI assessment Cardiac biomarkers, particularly troponins, have extensively been studied as a tool to evaluate chemotherapy cardiotoxicity. Elevations of serum cardiac troponin levels in asymptomatic patients after administration of anthracyclines predict deterioration of myocardial function (4). In contrast, very limited data exist on the role of troponins in the early detection of RT-induced cardiotoxicity. Hughes-Davies et al (10) quantified serum troponin T levels in 50 left-sided breast cancer patients before and after RT. They did not find any change in serum troponin T levels after RT. The patient population of this study differs from ours in that IM LNs were not irradiated, the total RT dose was lower, and none of the patients were previously treated with anthracyclines or trastuzumab. In our study, TnI levels were significantly elevated after RT compared with before RT in the left-sided group, whereas in the right-sided group, no change in TnI levels could be noticed. However, in all patients the TnI values remained below the quantification level. Therefore our findings could still be
1178 Erven et al.
International Journal of Radiation Oncology Biology Physics
coincidental, because the assay that was used is possibly not sensitive enough to guarantee the results of the TnI assessment below the quantification level.
Therefore we encourage the use of new techniques, such as respiratory gated RT, enabling reduced exposure of the anterior LV wall.
Limitations Interpatient variations in cardiotoxic risk may be caused by several patient- and treatment-related factors. Larger patient groups are needed to increase the power to detect the influence of other factors on post-RT strain/SR reductions. SRI should also be performed before the start of chemotherapy to be able to define the effect of chemotherapy on the observed strain/SR reductions. Dose-volume characteristics were calculated for 18 LV segments but not for the coronary arteries. Because the coronary arteries have a very small volume and no contrast was given for the planning computed tomography scan, exact contouring of the coronary arteries is problematic and prevents meaningful dosevolume statistics. The relevance of strain/SR reductions as a surrogate parameter for clinical long-term effects of RT on myocardial function needs further clarification. Radiation-induced cardiac disease may become clinically manifest only 10 years or more after treatment, so further investigation with long-term follow-up is needed to examine whether patients with larger reduction in strain/SR are more likely to have cardiac dysfunction. The immunoassay that was used for TnI measurement had a rather high quantification level and was therefore not ideal to use for detection of acute post-RT cardiotoxicity, which is expected to be only minor myocardial damage. To confirm the differences in TnI levels that were found after RT in the left-sided patient group and assess the value of troponins in the early detection of post-RT cardiotoxicity, a high-sensitivity assay should be used.
Conclusions SRI is a sensitive method to detect RT-induced changes in cardiac function that are not detectable by conventional echocardiographic measures. These subclinical reductions in cardiac function persist up to 14 months after RT. Further investigations should define whether SRI can be predictive for late cardiac morbidity. It might be an interesting tool to define patients at risk, who could possibly be saved by the early initiation of preventive measures. A critical dose level for cardiac dysfunction could not be proposed. Post-RT reduction in cardiac function was most prominent in the anterior LV wall of left-sided breast cancer patients.
References 1. Clarke M, Collins R, Darby S, et al. Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: An overview of the randomised trials. Lancet 2005;366:2087-2106. 2. Cuzick J, Stewart H, Rutqvist L, et al. Cause-specific mortality in long-term survivors of breast cancer who participated in trials of radiotherapy. J Clin Oncol 1994;12:447-453. 3. Gagliardi G, Constine LS, Moiseenko V, et al. Radiation dosevolume effects in the heart. Int J Radiat Oncol Biol Phys 2010; 76:S77-S85. 4. Altena R, Perik PJ, van Veldhuisen DJ, de Vries EG, Gietema JA. Cardiovascular toxicity caused by cancer treatment: Strategies for early detection. Lancet Oncol 2009;10:391-399. 5. Ganame J, Claus P, Eyskens B, et al. Acute cardiac functional and morphological changes after Anthracycline infusions in children. Am J Cardiol 2007;99:974-977. 6. Jurcut R, Wildiers H, Ganame J, et al. Strain rate imaging detects early cardiac effects of pegylated liposomal doxorubicin as adjuvant therapy in elderly patients with breast cancer. J Am Soc Echocardiogr 2008;21: 1283-1289. 7. Ho E, Brown A, Barrett P, et al. Subclinical anthracycline- and trastuzumab-induced cardiotoxicity in the long-term follow-up of asymptomatic breast cancer survivors: A speckle tracking echocardiographic study. Heart 2010;96:701-707. 8. Hare JL, Brown JK, Leano R, Jenkins C, Woodward N, Marwick TH. Use of myocardial deformation imaging to detect preclinical myocardial dysfunction before conventional measures in patients undergoing breast cancer treatment with trastuzumab. Am Heart J 2009;158:294-301. 9. Erven K, Jurcut R, Weltens C, et al. Acute radiation effects on cardiac function detected by strain rate imaging in breast cancer patients. Int J Radiat Oncol Biol Phys 2011;79:1444-1451. 10. Hughes-Davies L, Sacks D, Rescigno J, Howard S, Harris J. Serum cardiac troponin T levels during treatment of early-stage breast cancer. J Clin Oncol 1995;13:2582-2584. 11. Erven K, Petillion S, Weltens C, et al. Conformal locoregional breast irradiation with an oblique parasternal photon field technique. Med Dosim 2011;36:28-34. 12. Gagliardi G, Lax I, Soderstrom S, Gyenes G, Rutqvist LE. Prediction of excess risk of long-term cardiac mortality after radiotherapy of stage I breast cancer. Radiother Oncol 1998;46:63-71. 13. Kowalski M, Kukulski T, Jamal F, et al. Can natural strain and strain rate quantify regional myocardial deformation? A study in healthy subjects. Ultrasound Med Biol 2001;27:1087-1097.