Percutaneous revascularization in patients treated with thoracic radiation for cancer

Clinical Investigations

Percutaneous revascularization in patients treated with thoracic radiation for cancer Erin A. Fender, MD, a,1 Jackson J. Liang, DO, b,1 Terence T. Sio, MD, MS, c John M. Stulak, MD, d Ryan J. Lennon, MS, e Joshua P. Slusser, BS, e Jonathan B. Ashman, MD, PhD, c Robert C. Miller, MD, MS, c Joerg Herrmann, MD, a Abhiram Prasad, MD, a and Gurpreet S. Sandhu, MD, PhD a MN, PA, AZ, USA

Objectives To assess coronary revascularization outcomes in patients with previous thoracic radiation therapy (XRT). Background Previous chest radiation has been reported to adversely affect long term survival in patients with coronary disease treated with percutaneous coronary interventions (PCI).

Methods Retrospective, single center cohort study of patients previously treated with thoracic radiation and PCI. Patients were propensity matched against control patients without radiation undergoing revascularization during the same time period. Results

We identified 116 patients with radiation followed by PCI (XRT-PCI group) and 408 controls. Acute procedural complications were similar between groups. There were no differences in all-cause and cardiac mortality between groups (allcause mortality HR 1.31, P = .078; cardiac mortality 0.78, P = .49).

Conclusion Patients with prior thoracic radiation and coronary disease treated with PCI have similar procedural complications and long term mortality when compared to control subjects. (Am Heart J 2017;187:98-103.)

Cancer survival has substantially improved with advances in modern therapies. As the population of long term cancer survivors has increased, there has been a growing recognition of the delayed deleterious effects of cancer therapies. Thoracic external beam radiation therapy (XRT) is associated with later development of cardiovascular disease including coronary artery disease. 1,2 Historically, 50% of Hodgkin lymphoma survivors treated with mediastinal XRT developed cardiac disease, with cardiac disease accounting for 25% of the total mortality observed in this cohort. 2 In breast cancer survivors thoracic radiation has been associated with a 1.76-fold (95% CI: 1.34 to 2.31) higher risk of dying from cardiac disease as compared to controls. 1

From the aDepartment of Cardiovascular Diseases, Mayo Clinic, Rochester, MN, USA, b Division of Cardiovascular Disease, University of Pennsylvania, Philadelphia, PA, USA, c Department of Radiation Oncology, Mayo Clinic, Scottsdale, AZ, USA, dDivision of Cardiovascular Surgery, Mayo Clinic, Rochester, MN, USA, and eDivision of Biomedical Statistics and Informatics, Mayo Clinic, Rochester, MN, USA. 1 Both authors contributed equally to this publication. Disclosures: None of the authors has any disclosures or relevant relationships with industry. All authors have read and approved of the final article. Submitted August 2, 2016; accepted February 11, 2017. Reprint requests: Gurpreet S. Sandhu M.D., Mayo Clinic, 200 First Street SW, Rochester, MN 55905. E-mail: [email protected] 0002-8703 © 2017 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ahj.2017.02.014

The role of percutaneous coronary intervention (PCI) has been well established in the general population. However, outcomes in cancer survivors with previous thoracic radiation exposure remain unknown. We have previously shown that thoracic radiation prior to or following PCI does not increase the risk of late stent failure. 5 Given the durability of PCI in this cohort and the risk for subsequent progressive valve and pericardial disease which may necessitate a second sternotomy, catheter-based therapies may be the preferred method of revascularization, particularly when an acceptable internal mammary artery (IMA) conduit is not available. However, important questions remain including how radiation impacts procedural complications and long term survival following PCI. The aim of our study was to assess short term complications and long term survival following PCI in radiation treated patients.

Methods Study population This was a single center, retrospective cohort study of patients who received thoracic radiation and percutaneous coronary revascularization for CAD at the Mayo Clinic (Rochester, MN). 30,900 patients received thoracic radiation between 1971 and 2013. 16,578 patients underwent PCI from 1994 to 2013. After cross referencing the two patient lists, 155 patients were identified who underwent radiation prior to PCI. Next, patients

American Heart Journal Volume 187, Number 0

Fender et al 99

Table I. Location and type of cancer in the study cohorts Cancer type, N (%)

N = 116

Breast Esophageal Hodgkin lymphoma Non-Hodgkin's lymphoma Lung Mediastinum Thoracic spinal cord Stomach Thyroid N 1 type of cancer

59 (51) 13 (11) 11 (9) 3 (3) 19 (16) 3 (3) 1 (1) 3 (3) 4 (3) 1 (1)

were excluded if the radiation field did not involve the heart or if they were treated with palliative intent radiation. Propensity matched controls were selected from patients treated with PCI over the same time period but without previous thoracic radiation exposure. Propensity score matching was performed on multiple clinical and procedural factors which are available in Supplemental Table I.

Radiation therapy All patients had a diagnosis of cancer (biopsy proven or radiographically early stage non-small cell lung cancer), and were treated with curative intent radiation including standard or intensity-modulated radiotherapy (Table I). Over 90% of patients received radiation for lymphoma, breast, lung, or esophageal cancer. A radiation oncologist reviewed all computed tomography radiation simulation plans and confirmed cardiac involvement for each patient. PCI database The Mayo Clinic PCI registry includes demographic, angiographic, and procedural data for patients treated with PCI at our institution. In-hospital events are recorded, and follow up data including all-cause and cardiac mortality is obtained via telephone contact by a research coordinator using a questionnaire at 6 months, 1 year, and then annually. Patients with angioplasty and/or stenting from 1994–2013 (including bare metal and drug eluting stents) were included. All patients were prescribed dual antiplatelet therapy (1 month for bare metal stents, 12 months for drug eluting stents), and lifelong aspirin therapy was recommended. Outcomes The primary outcome was all-cause mortality. Secondary outcomes included cardiac and non-cardiac mortality and procedural characteristics including complications, the number and location of diseased vessels, number of vessels intervened on, number and type of stents deployed, and angiographic characteristics.

Statistical methods Continuous variables are summarized as mean ± SD. Discrete variables are summarized as a frequency (percentage). A propensity score was created using logistic regression (Supplemental Table I). Optimal variable matching was used with up to 4 reference subjects matched to each radiation patient. 6 Reference subjects were chosen according to age (within 5 years), gender, date of revascularization (within 2 years), and propensity score (within 1/4 the SD of the propensity score distribution). When calculating descriptive statistics, weighting was used to account for the different number of referent subjects matched to each radiation case. Conditional logistic regression was used to test the difference between radiation subjects and their matched controls. Kaplan–Meier methods were used to estimate all-cause mortality. Competing-risks methods were used to estimate the incidence of cause-specific mortality. To test differences in survival, Cox proportional hazards models were applied with a frailty term for each set of matched subjects. Follow-up time was measured starting from the date of revascularization to date of death or censor. All analyses were performed using SAS version 9.3 or higher (SAS Inc, Cary, NC). All hypotheses tests were 2-sided with a 0.05 significance level. This study was conducted with institutional review board approval. No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, the study analysis, and drafting and editing of the paper.

Results Percutaneous coronary intervention A total of 116 cancer survivors previously treated with thoracic radiation underwent PCI (XRT-PCI patients). The median radiation to PCI interval was 5.6 years (interquartile range [IQR] 1.1, 12.3 years). We identified 408 propensity matched control patients. Baseline and demographic characteristics for can be found in Table II, while angiographic and procedural characteristics are outlined in Table III. Patients were well matched on both clinical and angiographic variables. Compared with controls, XRT-PCI patients had less hypertension (73% vs 83%, P = .025) and more often had an ejection fraction ≤40% (18% vs 13%, P = .043). Previous radiation did not result in a higher rate of PCI complications (Table III). After a median follow-up of 6.3 years after PCI (IQR 4.0, 9.8 years), there was no difference in all-cause, cardiac, or non-cardiac mortality between XRT-PCI patients and controls (Figure A-B). Fifty-six deaths occurred in the XRT-PCI group (including 12 due to cardiac causes), and 159 deaths occurred in the controls (including 44 cardiac deaths).

American Heart Journal Month 2017

100 Fender et al

Table II. Baseline clinical characteristics at the time of percutaneous coronary revascularization for radiation treated patients (XRT-PCI) and matched controls Variable, N (%) Age, years Male gender Body mass index Diabetes mellitus Hypertension Hyperlipidemia Former or current smoker Prior myocardial infarction Prior PCI Prior CABG Congestive heart failure Peripheral vascular disease History of stroke or transient ischemic accident Moderate or severe renal disease Chronic obstructive pulmonary disease Left ventricular ejection fraction

XRT-PCI (N = 116)

Controls (N = 408)

70.0 ± 10.0 43 (37) 28.5 ± 5.4 29 (25) 83 (73) 84 (74) 75 (67) 27 (23) 18 (16) 12 (10) 22 (20) 15 (13) 16 (14)

70.3 ± 10.0 151 (37) 29.0 ± 5.6 113 (29) 320 (83) 298 (77) 257 (65) 79 (20) 64 (16) 43 (11) 86 (22) 79 (20) 51 (13)

4 (3)

28 (7)

Controls (N = 408)

1 (1)

9 (2)

32 (30)

119 (31)

40 (38)

144 (38)

33 (31)

111 (29)

79 (71) 50 (50) 40 (37) 21 (20) 19 (19) 1.3 ± 0.6

257 (67) 195 (51) 112 (29) 57 (15) 87 (25) 1.3 ± 0.6

•1 •2 •3

99 (87)

352 (87)

15 (13)

51 (13)

0 (0)

2 (0)

1.3 ± 0.8 53 (46) 65 (56)

1.3 ± 0.8 209 (52) 217 (53)

• Left main coronary artery • Left anterior descending coronary artery • Right coronary artery • Left circumflex artery • Vein graft

5 (4)

14 (3)

.89

44 (39)

164 (41)

.99

42 (37)

164 (41)

.31

36 (32)

115 (29)

.48

1 (1)

2 (0)

.90

• • • • • • • •

1 (1)

19(5)

.037

4(3)

13(3)

.62

1(1)

6(2)

.56

4(3)

11(3)

0

0

0

3(1)

.26

2(2)

10(2)

.54

1(1)

6(2)

.5

Variable, N (%)

.19 1.00 .37 .41 .025 .42 .66 .56 .94 .91 .46 .10 .59

Number of diseased vessels

.23

86 (21)

• ≤40% • N40% • Unknown

21 (18)

52 (13)

46 (40)

207 (51)

49 (42)

149 (36)

• Elective • Urgent • Emergent

30 (26)

135 (33)

53 (46)

184 (45)

33 (28)

89 (22)

5 (4) 2 (2)

23 (6) 8 (2)

.72 .71

61 (53)

192 (47)

.28

Pre-procedural shock Prophylactic intra-aortic balloon pump Canadian Heart Class (III, IV, or V)

XRT-PCI (N = 116)

P

20 (17)

.21

.043

Urgency of PCI

Table III. Angiographic and procedural characteristics at the time of percutaneous coronary revascularization for radiation treated patients (XRT-PCI) versus controls

.06

Discussion We observed no difference in all-cause or cardiac mortality in the study group versus controls. Additionally, radiation exposure did not increase periprocedural adverse events among the PCI patients. Recently we found no differences in the rate of stent failure among PCI-treated patients with radiation exposure compared to the general population. 5 In the current study, we observed no difference in PCI procedural characteristics or in long term mortality. We did note in Figure A a non-significant trend towards increased all-cause mortality in the XRT-PCI population after 2 years, however this was not significant despite adequate powering. Additionally, in Figure B it becomes clear that if a difference in all-cause mortality really did exist, it was likely driven by a non-significant trend towards increased non-cardiac mortality in the study group. When taken together, this study suggests PCI may perform equally well in controls and radiation patients.

• • • •

0 1 2 3

Multi vessel disease Type C lesion Thrombus in any lesion Bifurcation in any lesion TIMI 0 or 1 pre-PCI in any lesion Number of segments treated Total number of vessels treated

Total number of stents placed Number of drug-eluting stents Intra-procedural GPIIb/IIIa use Vessel(s) intervened on with PCI

P .79

.50 .71 .14 .21 .35 .91 .95

.86 .31 .54

Procedural complications

Myocardial infarct Branch occlusion Cardiac arrest Shock Tamponade Stroke Renal failure Death

.67 1.00

Our finding of similar long term mortality in radiation and control patients following PCI differs from those recently published by Reed et al. 7 Reed examined mortality in 157 PCI patients with previous chest radiation and observed 37 cardiac deaths in the radiation patients and 32 cardiac deaths in control patients (HR 1.7, 95% CI 1.06–2.89, P = .03). Patients were matched on age, sex, lesion artery, and type of PCI. However important differences in the baseline characteristics were present including twice the rate of atrial fibrillation (27% vs 11%, P b .001), COPD (24% vs 13%, P = .026), smoking (16% vs 7%, P = .02), and presentation with non-ST elevation myocardial infarction (13% vs 8%, P = .15) in the radiation patients. Radiation patients were also twice as likely to have more complex lesions coded as B2/C (71% vs 36%, P b .001). In the current study patients were propensity matched

American Heart Journal Volume 187, Number 0

Figure

Fender et al 101

American Heart Journal Month 2017

102 Fender et al

using both clinical and lesion characteristics (Supplemental Table I), including the urgency of PCI, and were well matched. It is possible imbalances in baseline comorbidities in the Reed study account for the higher mortality observed in the radiation cohort. Characterizing the performance of PCI in patients with previous radiation is critical, particularly as increased mortality following surgical revascularization has been reported. 3 Furthermore, multiple studies have documented an adverse effect of radiation on internal mammary artery (IMA) integrity. 8,9 This has significant clinical implications as the IMA provides the greatest long term survival benefit in patients undergoing CABG. 10-12 In radiation patients without a viable IMA, PCI may prove superior due to fewer procedural complications and a shorter recovery time relative to bypass surgery. Furthermore, many radiation patients go on to develop significant valvular or pericardial disease which may necessitate cardiac surgery at a later date. By avoiding an early sternotomy for coronary revascularization, the risk of subsequent cardiac surgeries for valvular or pericardial disease may be reduced. The IMAs are typically not injected during preoperative angiography as they are nearly always patent in patients without previous chest radiation. However, in patients with thoracic radiation exposure routine preoperative IMA angiography should be considered to assess patency and thereby help guide the revascularization strategy.

Limitations This research has several important limitations; primary among them is the exclusion of patients who were treated with radiation outside of our institution. We sought to include only patients in whom cardiac involvement within the radiation field could be confirmed by a radiation oncologist and therefore patients who received radiation elsewhere were excluded. This resulted in higher quality of data but a smaller sample size. An additional limitation is the heterogeneity in the amount and type of radiation delivered to individual patients. Before the late 1980s, radiation doses of 35–41 Gy were administered to Hodgkin patients. Modern regimens now employ 20–30 Gy delivered to smaller volumes. 13 Unfortunately, doses of 45–50 Gy are still employed in locally advanced breast cancer, although often through beams passing tangentially through the distal ventricles. 14 These differences in XRT exposure by type of cancer and year of treatment may confound our results as the incidence and severity of radiation heart disease is proportionate to dose exposure. 4 However, the types of cancers in this cohort are similar to those

commonly observed in clinical practice to cause significant radiation heart disease, with greater than 90% of patients carrying a diagnosis of lymphoma, breast, lung, or esophageal cancer (Table I).

Conclusions We demonstrate percutaneous revascularization in patients with prior thoracic XRT exposure can be performed with a safety profile comparable to propensity matched control patients. Additionally, long term survival was not significantly different between study subjects and controls. Previous research has suggested that XRT may compromise the IMAs, and therefore in patient in whom coronary bypass surgery is being considered, preoperative IMA angiography should be considered. Given the favorable performance of PCI in this cohort, it is reasonable to consider percutaneous revascularization in patients with radiation heart disease who do not have a patent IMA on preoperative angiography or without left main or left anterior descending artery coronary disease. Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.ahj.2017.02.014.

Acknowledgments All individuals who contributed significantly to this project have been listed as authors and all have authors have approved the final article. None of the authors has any financial disclosures or relationships with industry.

References 1. Bouillon K, Haddy N, Delaloge S, et al. Long-term cardiovascular mortality after radiotherapy for breast cancer. J Am Coll Cardiol 2011;57:445-52. 2. Lee CK, Aeppli D, Nierengarten ME. The need for long-term surveillance for patients treated with curative radiotherapy for Hodgkin's disease: University of Minnesota experience. Int J Radiat Oncol Biol Phys 2000;48:169-79. 3. Wu W, Masri A, Popovic ZB, et al. Long-term survival of patients with radiation heart disease undergoing cardiac surgery: a cohort study. Circulation 2013;127:1476-85. 4. Chang AS, Smedira NG, Chang CL, et al. Cardiac surgery after mediastinal radiation: extent of exposure influences outcome. J Thorac Cardiovasc Surg 2007;133:404-13. 5. Liang JJ, Sio TT, Slusser JP, et al. Outcomes after percutaneous coronary intervention with stents in patients treated with thoracic external beam radiation for cancer. JACC Cardiovasc Interv 2014;7: 1412-20. 6. Ming K, Rosenbaum PR. Substantial gains in bias reduction from matching with a variable number of controls. Biometrics 2000;56:118-24.

A, Kaplan–Meier survival curve from the time of percutaneous coronary revascularization (PCI) in patients treated with radiation versus matched controls. Hazard ratio 1.31; 95% CI 0.97–1.76 P = .078. B, Incidence of cardiac death and non-cardiac death for radiation patients treated with percutaneous coronary intervention versus matched controls. Cardiac death hazard ratio 0.78 (95% CI 0.376–1.600; P = .49). Non-cardiac death hazard ratio 1.35 (95% CI 0.932–1.947; P = .11).

American Heart Journal Volume 187, Number 0

7. Reed GW, Masri A, Griffin BP, et al. Long-Term Mortality in Patients With Radiation-Associated Coronary Artery Disease Treated With Percutaneous Coronary Intervention. Circ Cardiovasc Interv 2016;9. 8. Brown ML, Schaff HV, Sundt TM. Conduit choice for coronary artery bypass grafting after mediastinal radiation. J Thorac Cardiovasc Surg 2008;136:1167-71. 9. van Son JA, Noyez L, van Asten WN. Use of internal mammary artery in myocardial revascularization after mediastinal irradiation. J Thorac Cardiovasc Surg 1992;104:1539-44. 10. Cameron A, Davis KB, Green GE, et al. Clinical implications of internal mammary artery bypass grafts: the Coronary Artery Surgery Study experience. Circulation 1988;77: 815-9.

Fender et al 103

11. Loop FD, Lytle BW, Cosgrove DM, et al. Influence of the internal-mammary-artery graft on 10-year survival and other cardiac events. N Engl J Med 1986;314:1-6. 12. Cameron AA, Green GE, Brogno DA, et al. Internal thoracic artery grafts: 20-year clinical follow-up. J Am Coll Cardiol 1995;25:188-92. 13. Eichenauer DA, Engert A, Andre M, et al. Hodgkin's lymphoma: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2014;25(Suppl 3):iii70-5. 14. Hojris I, Overgaard M, Christensen JJ, et al. Morbidity and mortality of ischaemic heart disease in high-risk breast-cancer patients after adjuvant postmastectomy systemic treatment with or without radiotherapy: analysis of DBCG 82b and 82c randomised trials. Radiotherapy Committee of the Danish Breast Cancer Cooperative Group. Lancet 1999;354:1425-30.