Thoracic Radiation Therapy and Suitability of Internal Thoracic Arteries for Myocardial Revascularization

Thoracic Radiation Therapy and Suitability of Internal Thoracic Arteries for Myocardial Revascularization* Giuseppe Nasso, MD; Carlo Canosa, MD; Carlo Maria De Filippo, MD; Pietro Mondugno, MD; Amedeo Anselmi, MD; Mario Gaudino, MD; and Francesco Alessandrini, MD

Introduction and background: Myocardial revascularization using internal thoracic arteries (ITAs) has been associated with superior clinical outcome. This study addresses the question of whether internal mammary arteries are unsuitable for grafting due to radiation-based damage in patients with history of thoracic radiation therapy. We review our experience in this subset of surgical candidates. Patients and methods: Forty-nine patients undergoing elective coronary artery bypass grafting with use of at least one ITA were enrolled and matched to 49 comparable nonirradiated individuals by propensity scoring system. Preoperative and postoperative data were collected and compared. A 18-month clinical follow-up was performed. Results: Intraoperative mammary artery flow was assessed by transthoracic Doppler echocardiography probe, and there was no significant difference between irradiated and nonirradiated individuals (36 ⴞ 8.3 mL/min vs 39 ⴞ 7.2 mL/min, p ⴝ 0.15). The two study groups were also comparable in terms of survival (overall mortality, 2%) and recurrence of angina and perfusion defect at control stress perfusion nuclear scan (p ⴝ 0.99 and p ⴝ 0.77, respectively). One arterial graft showed stenosis at postoperative angiography. The dose of radiation therapy administered did not correlate with graft flow values after anastomosis. Conclusion: Our data suggest that the use of a monolateral or bilateral ITA is not associated with early graft failure in patients with history of chest/mediastinal irradiation. Skeletonization harvesting technique might be recommended because of the frequent presence of fibrous tissue around the in situ vessel. Arterial graft stenosis in thorax-irradiated patients should be attributed to the primary atherosclerotic disease rather than to irradiation itself. (CHEST 2005; 128:1587–1592) Key words: coronary artery bypass grafting; internal thoracic artery; radiotherapy Abbreviations: CABG ⫽ coronary artery bypass grafting; ITA ⫽ internal thoracic artery; LAD ⫽ left anterior descending artery; LITA ⫽ left internal thoracic artery; NYHA ⫽ New York Heart Association; RITA ⫽ right internal thoracic artery; RT ⫽ radiation therapy

internal thoracic artery (ITA) is diffusely T headopted for myocardial revascularization. Arte-

rial grafting is an advisable approach even in patients with a major atherosclerotic burden.1 Evidence in the literature supports the concept that arterial grafts have better patency rates over time and are associated with superior patient outcome at follow-up.2,3 Data in the literature concerning the outcome of ITA grafts in patients with a history of thoracic *From the Department of Cardiac Surgery, Center for High Technology Research and Education in Biomedical Sciences, Universita` Cattolica del Sacro Cuore, Campobasso, Italy. Manuscript received February 5, 2005; revision accepted April 18, 2005. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Giuseppe Nasso, MD, via Acaia 24, 00183 Roma, Italy; e-mail: [email protected] www.chestjournal.org

radiation therapy (RT) are, however, conflicting. Some reports4,5 suggest that RT may render ITAs unsuitable for grafting after lying within the radiation field. Meanwhile, larger systematic studies6,7 found no adverse effects on patient outcome. Irradiation is a potential trigger for intimal damage,8,9 which may lead either to arterial occlusion by accelerated atherosclerosis10 or to disruption in several arterial districts.11,12 While coronary artery lesions requiring surgical revascularization may be more frequent in patients who undergo thoracic RT, surgeons often avoid the use of ITAs in these individuals, and a saphenous vein or alternative arterial conduit is preferred. The aim of this study was to revise our experience with short-term and mid-term outcomes of patients who underwent ITA grafting with and without previous RT for breast or mediastinal canCHEST / 128 / 3 / SEPTEMBER, 2005

1587

cer, in order to provide evidence about the clinical results of irradiated ITA grafts. Materials and Methods Patient Selection From January 1998 to May 2003, a total of 678 patients underwent elective coronary artery bypass grafting (CABG). Among these, 63 patients (9.3%) had history of thoracic RT for cancer. Study individuals were preliminarily selected according to the following criteria: (1) isolated elective myocardial revascularization; reoperative cardiac procedures were excluded; (2) on-pump CABG with use of at least a monolateral ITA, while all procedures with employment of any arterial conduit other than an ITA (gastroepiploic or radial artery) were excluded; (3) history of surgery for breast tumor (either tumorectomy or quadrantectomy or mastectomy followed by axillary lymph nodes excision, when indicated) or history of surgery for thymoma or history of Hodgkin or non-Hodgkin lymphoma with mediastinal “bulky” localization; all patients were radically treated and were relapse free at the time of cardiac surgery; and (4) thoracic RT (at least 30 Gy total dose) administered within 24 to 6 months before the intervention using a field involving at least the mediastinum or the left hemithorax. A second group of individuals who satisfied both criteria 1, 2, and 3 but had not received RT according to oncologist indications was selected. Screened patients had been addressed to our center by territorial oncology units. Selection was retrospectively made among the records included in our electronic database. The two groups underwent propensity scoring matching. Preoperative and postprocedural features were collected and compared. Study End Points Study end points were as follows: (1) intraoperative and postoperative evaluation of ITA grafts; (2) angina recurrence at 18 months and perfusion defect at control perfusion nuclear scan within 12 months; and (3) global clinical results.

one LITA-to-LAD or a LITA-to-diagonal branch anastomosis or a Y-shaped graft involving the LITA. The great saphenous vein was variably adopted for non-LAD revascularization in all further cases. After conduits harvesting, extracorporeal circulation was instituted by right atrial and ascending aorta cannulation. Myocardial protection was accomplished by antegrade warm blood multidose cardioplegia. Arterial conduits were divided only after a clotting time ⬎ 400 s to avoid intraluminal thrombosis; they were tested for 15 s to verify the flow, and all arteries were carefully inspected to detect intraluminal thrombosis or luminal dissection. Manipulation was done as gently as well. Immediately before performance of anastomosis, a papaverin solution was dropped on conduits in order to reduce vasospasm. After weaning from cardiopulmonary bypass, the flow in each arterial graft was measured by Doppler probe of the graft immediately after a 80 mm Hg systolic pressure had been reached. Segments of the ITA discarded during surgery were collected and stored in buffered paraformaldehyde for production of pathology sections. These were reviewed by two skilled pathologists in a blinded way. Statistical Analysis Continuous variables are expressed as mean ⫾ SD. A twotailed Student t testing for continuous variables and ␹2 test for discrete variables were used for group comparisons. After preliminary selection, the population was made by 63 individuals with history of RT fulfilling all inclusion criteria (group A) and by 49 patients fulfilling only criteria 1, 2, and 3 (group B). All these patients were subjected to logistic regression to model the probability of being assigned to group A or group B, with respect to baseline features, and the propensity score was calculated for each of them. For every group B patient, matching patients with the closest score were selected from the larger pool of group A patients (maximum allowable difference, 0.1). Fourteen group A patients were excluded, as no matching patient could be found among group B patients. An ␣ level of 0.05 was adopted. Among group A individuals, we tested total RT dose and LITA flow for correlation by Spearman R test, being LITA flow elimination was the response variable. Statistical analysis was performed using software (Statistical Package for Social Sciences, Release 11.0 for Windows; SPSS; Chicago, IL); propensity scoring analysis was performed using software (Release 8.2; SAS; Cary, NC).

Preoperative Evaluation Chest radiography, transthoracic echocardiography, and coronary angiography with cineventriculography were performed in all patients. Bilateral ITA angiography was additionally included for evaluation of patency. Surgical Strategy No major changes were introduced in our surgical and anesthesiology protocols during the study period. Following median sternotomy, the left ITA (LITA) and/or right ITA (RITA) were harvested skeletonized before heparinization. Skeletonization is routinely performed in our center mainly in diabetic and obese patients, as in our experience in this subgroup it can be protective against development of mediastinitis. The LITA was used to revascularize the left anterior descending artery (LAD) artery or the LAD and the diagonal artery branch by sequential grafting. The RITA was used to revascularize the right coronary artery or the posterior descending artery branch. In case of proximal LAD and circumflex artery lesion without involvement of the right coronary system, an Y graft was performed using a LITA circumflex graft and an isolated RITA segment to revascularize the LAD or a diagonal branch. All study individuals had at least 1588

Follow-up Follow-up lasted 18 months after discharge and was 100% complete. All surviving patients were scheduled and contacted for follow-up visits at 1, 6, 12, and 18 months, and additional visits when clinically indicated, including clinical examination, ECG, transthoracic cardiac ultrasonography, and a functional test by myocardial stress perfusion nuclear scan within 12 months from the procedure. ITA graft angiography was performed when noninvasive test results for myocardial ischemia were positive.

Results Early Results Both groups included 49 patients. Baseline features are summarized in Table 1. Overall mean age was 57 ⫾ 6.4 years (range, 48 to 81 years). There were no cases of occluded in situ ITA at preoperative angiography and one case of stringed right ITA. The propensity scoring system identified two groups of Clinical Investigations

Table 1—Preoperative Features*

Variables Males/female gender Diabetes Obesity Systemic hypertension Previous myocardial infarction Left ventricular ejection fraction ⱕ 40% NYHA III or IV Oncology history Breast tumor Thymoma Hodgkin lymphoma RT to cardiac surgery time, d† Associated chemotherapy

Group Group p Value A B (n ⫽ 49) (n ⫽ 49) Matched All 8/41 9 7 25 19

6/43 9 6 24 21

0.77 0.99 0.99 0.83

0.26 0.90 0.95 0.04 0.05

6

8

0.77

0.28

31

34

0.66

0.03

41 5 3 395 ⫾ 84

42 4 2

0.99 0.99 0.99

0.006 0.46 0.58

14

15

0.99

0.036

*Data are presented as No. unless otherwise indicated. †Mean ⫾ SD.

patients highly homogenous in terms of preoperative characteristics, and no significant differences in terms of sex, cardiovascular risk factors, New York Heart Association (NYHA) class, and oncology history were detected between the two groups. Female patients were prevalent in the overall study population, as most patients enrolled had history of breast cancer. Only a minority of individuals had history of non-breast tumor, including thymoma (10.2% in group A and 8.2% in group B) or Hodgkin lymphoma (6.1% and 4.1%, respectively). Mean interval between RT and cardiac surgery in group A was 395 ⫾ 84 days (range, 193 to 594 days). Mean radiation total dose administered per group A patient was 58 ⫾ 10 Gy, ranging from 30 Gy (mediastinal Hodgkin lymphoma) to 70 Gy (breast cancer staged as T4 according to TNM system). This total amount had been delivered in several administrations (mean, 21.3 Gy per patient). Group B patients had been successfully treated by surgery alone or by surgery plus chemotherapy or by chemotherapy alone (lymphomas).

Details of grafting strategy are given in Table 2. Total operative time was 231 ⫾ 47 min in group A vs 219 ⫾ 55 min in group B (p ⫽ 0.08). Cardiopulmonary bypass time was 62 ⫾ 20 min in group A and 65 ⫾ 16 min in group B (p ⫽ 0.13). During harvesting of mammary arteries, mild adhesions were encountered in the immediate substernal planes and around the artery itself when homolateral to the irradiated area. At gross inspection, no signs of intraluminal dissection were evidenced, while artery flow, measured during harvest by Doppler analysis of the graft, was 36 ⫾ 8.3 mL/min in group A vs 39 ⫾ 7.2 mL/min in group B (p ⫽ 0.15). No inverse correlation could be evidenced between total administered RT dose and graft flow (r ⫽ ⫺ 0.09, p ⫽ 0.68) [Fig 1]. Table 3 reports periprocedural results. There was one in-hospital death in group A (1% of the study population) due to perioperative stroke, and one nonfatal myocardial infarction. There was no significant difference in terms of ICU stay (p ⫽ 0.52) and hospital stay (p ⫽ 0.20), with 84% of patients discharged within 8 days. Reoperation for bleeding was necessary in one group A patient; the source of bleeding was not related to any arterial graft. Deep sternal wound infection was diagnosed on the basis of wound dehiscence with sternal instability associated to fever or hypothermia and to positive microbiological culture findings and occurred in two patients; both belonged to group B and had undergone bilateral ITA grafting (p ⫽ 0.49). Transient episodes of atrial fibrillation were observed in 13 individuals. At examination of pathology sections of ITA segments, no differences were evident between irradiated and nonirradiated arteries in terms of intimal and medial thickening. Figure 2 displays two sections from a patient belonging to group A and one from group B. Displayed sections are representative examples of the morphology encountered in specimens obtained from group A (n ⫽ 31) and group B (n ⫽ 37).

Table 2—Grafting Management Details* Variables LITA to LAD LITA to LAD to diagonal artery branch (sequential grafting) RITA to posterior descending artery RITA to right coronary artery Y graft Great saphenous vein to circumflex *Data are presented as No. www.chestjournal.org

Group A

Group B

p Value

31 9

35 6

0.51 0.57

15 10 9 12

16 12 8 15

0.99 0.80 0.99 0.65

Figure 1. Left ITA flow and RT total dose in each group A patient. CHEST / 128 / 3 / SEPTEMBER, 2005

1589

Table 3—Periprocedural Results* Variables

Group A

Table 4 —Eighteen-Month Follow-up Results

Group B p Value

In-hospital death 1 0 ICU stay, d 2.3 ⫾ 1.3 2.8 ⫾ 1.7 Revision for bleeding 1 0 Mediastinitis 0 2 Perioperative myocardial infarction 1 0 LITA graft flow, mL/min 36 ⫾ 8.3 39 ⫾ 7.2

0.99 0.52 0.99 0.49 0.99 0.15

Variables

Group A

Group B

p Value

Eighteen-month survival, % Angina recurrence, No. Perfusion defect, No.*

98 1 4

98 1 2

0.67

*At control stress myocardial nuclear scan.

*Data are presented as No. or mean ⫾ SD.

Late Results Follow-up was 100% complete. There was one late death in group B due to refractory heart failure, with an overall mortality rate of 2% in the study population (Table 4). At the end of the follow-up, 87 patients (88.8%) were in NYHA functional class I (42 in group A and 45 in group B, p ⫽ 0.52), while 11 patients were in class II (6 in group A and 5 in group B, p ⫽ 0.99). At control stress myocardial scan, defect of perfusion was evidenced in six individuals;

among these, two patients had presented with angina (one patient in group A and one in group B). Patients with positive ischemia test results underwent angiography, which revealed irregular involvement of LITA in one patient in group A with history of mediastinal lymphoma (2% of the LITA grafts in the group), and subocclusive stenosis of a saphenous vein graft in one patient in group B (5.3% of the vein grafts in the group). Both these patients were affected by severe hypercholesterolemia, diabetes mellitus, and systemic hypertension. Late percutaneous graft angioplasty was successfully adopted in both individuals, and the postprocedural course was uneventful; symptoms almost disappeared after percutaneous revascularization of grafts and institution of tailored medical therapy. Concerning the remaining four patients with positive nuclear scan findings, angiography failed to reveal any involvement of arterial grafts. Among these, three patients were mildly symptomatic (NYHA class II) and could be treated medically. Comment

Figure 2. Top, A: Pathology section of an ITA obtained from a patient who underwent RT for mediastinal lymphoma. Bottom, B: Pathology section of an ITA obtained from a patient belonging to the control group (hematoxylin-eosin, original ⫻ 20). 1590

Surgeons might feel discouraged from performing myocardial revascularization with ITAs in patients with history of thoracic RT. Irradiation may act as a trigger of intimal damage following intimal denudation and thrombus formation, and of media damage that gives rise to fibrosis and cell proliferation.13 These mechanisms, which are the pathogenetic leitmotiv of early failure of venous grafts and of coronary atherosclerosis in heart transplant,10 in the irradiated arteries may lead to accelerated atherosclerosis and impaired patency of grafts.14 Thoracic RT may also determine the presence of scar tissue around mammary arteries and increase atherosclerotic burden of coronary arteries. Furthermore, no studies are available concerning the different extent of RT-related damage in muscular or elastic arteries, as these conduits show different vasoreactive profiles and muscular cells proliferation is among the determinant events of occlusive processes. Anyway, most clinical reports4,5 supporting the concept that thoracic RT impairs the suitability of ITA as a graft are indeed anecdotal and probably do not provide reliable evidence on this topic. Our study tries to Clinical Investigations

overcome the limits inherent its retrospective nature by the enrollment of a study population larger than others reported in literature6,7 and by the adoption of propensity scoring in order to boost comparability of groups. We aimed at evaluating the midterm clinical outcome of irradiated patients submitted to CABG with mammary artery vs comparable nonirradiated individuals, and the patency rates of this artery in the same settings. In our study cohort, at intraoperative gross examination, no mammary artery showed signs of dissection or of intraluminal thrombosis. We noticed mild sclerosis of the wall of internal thoracic veins homolateral to harvested ITAs when included in the irradiation field, and adhesions with some calcific spot around the ITA itself. Despite the need for accurate dissection of such adhesions during skeletonization, operative time was not significantly longer in group A (p ⫽ 0.08). ITAs had similar blood flow values before chest closure. At follow-up, history of administration of radiotherapy seems not to influence the 18-month outcome after CABG using either bilateral ITAs or monolateral LITAs, as we observed comparable rates of survival, angina-free survival, and perfusion defect at noninvasive testing. Among the 14 cases of defect of perfusion, only in 1 case was there angiographic evidence of arterial graft involvement. In the present experience, only connective tissue surrounding the mammary artery seems to undergo “sclerotic” remodeling after RT, while the artery wall seems to be not significantly affected. Adhesions surrounding the ITA before the intervention may account for a “restrictive” pattern of the artery, with reduction of vessel wall compliance and flow; skeletonization probably represents the best option for restoring optimal mechanical properties of mammary arteries. However, the presence of tight adhesions around the thin-walled artery in some irradiated patients might render skeletonization hazardous. Intraoperatively surgeons should therefore take into account and balance both these factors in the choice of pedicled or skeletonized technique. Therefore, if at preoperative ITA angiography there is no evidence of focal atherosclerotic plaques, surgeons should keep much more confident in the vessel if skeletonization harvest is programmed. Further evidences are anyway necessary to definitely clarify this issue. Graft flow evaluated in the left ITA in group A showed no correlation with the total RT dose administered; ie, our data do not suggest a dose-dependent effect of radiation on graft flow after skeletonization and anastomosis (Fig 1). These results agree with the findings by Gansera and coworkers.6 All cases of mediastinitis diagnosed in our series occurred in www.chestjournal.org

group B in two diabetic individuals who both underwent bilateral ITA grafting; these results is consistent with others in the literature.15-17 Despite the limit of incomplete angiographic follow-up, we found that arterial revascularization using either monolateral or bilateral mammary arteries has optimal results even in patients who had received thoracic irradiation, in terms of mortality and clinical outcome at 18 months. The absence of symptoms and negative stress ischemia test findings suggest that arterial grafts are functional even in patients who underwent no control angiography, despite morphologic evaluation of grafts performed in further investigations to remove all doubts. Further data will be gained about the suitability of the gastroepiploic artery in patients who had received upper-abdominal RT.18 In conclusion, our evidence suggests that surgeons should be confident with ITAs even in patients with history of thoracic or mediastinal RT. Our data suggest that ITA graft flow is comparable between irradiated and nonirradiated individuals, which might have a favorable impact on angina-free survival of these patients given the recognized superiority of arterial grafts at long-term follow-up. Such a finding awaits confirmation by long-term follow-up studies, as the present report is limited by the analysis of the sole intraoperative and 1-year results. The good functional performance of irradiated ITAs documented here encourages their use and investigation of overall angiographic outcome. Also the use of bilateral ITA grafting should not be discouraged in irradiated patients, as it is likely to determine a better prognosis,19 and irradiation seems not in our cohort to represent an additional risk factor to others reported in literature16,17 for mortality and postoperative morbidity. The relatively few number of patients included in our study does not allow a formal analysis of risk factors and of survival estimates. Although further studies are needed to confirm this concept, ITA grafts in irradiated patients should be considered subjected to the common atherosclerotic risk factors (hypercholesterolemia, diabetes) without clinically relevant additional risks. If arterial graft involvement occurs, this should be ascribed to the patient’s primary atherosclerotic disease. These findings are to be confirmed in arteries with larger muscular component adopted for myocardial revascularization. References 1 Kay HR, Korns ME, Flemma RJ, et al. Atherosclerosis of the internal mammary artery. Ann Thorac Surg 1976; 21:504 –507 2 Chavanon O, Durand M, Hacini R, et al. Coronary artery bypass grafting with left internal mammary artery and right gastroepiploic artery, with and without bypass. Ann Thorac Surg 2002; 73:499 –504 CHEST / 128 / 3 / SEPTEMBER, 2005

1591

3 Wandschneider W, Wurning P, Redtenbacher S, et al. Arterial grafts for coronary artery surgery. Cardiovasc Surg 1995; 3:525–527 4 Schulman HE, Korr KS, Myers TJ. Left internal thoracic artery graft occlusion following mediastinal radiation therapy. Chest 1994; 105:1881–1882 5 Renner SM, Massel D, Moon BC. Mediastinal irradiation: a risk factor for atherosclerosis of the internal thoracic arteries. Can J Cardiol 1999; 15:597– 600 6 Gansera B, Haschemi A, Angelis I, et al. Cardiac surgery in patients with previous carcinoma of the breast and mediastinal irradiation: is the internal thoracic artery graft obsolete? Thorac Cardiovasc Surg 1999; 47:376 –380 7 Van Son JA, Novez L, van Asten WN. Use of internal mammary artery in myocardial revascularization after mediastinal irradiation. J Thorac Cardiovasc Surg 1992; 104:1539 – 1544 8 Bartels C, Erasmi A, Sayk F, et al. Prophylactic ␥ radiation of unaffected vein grafts failed to prevent vein graft disease in a chronic hypercholesterolemic porcine model. Eur J Cardiothorac Surg 2003; 24:92–97 9 Teirstein PS, Massullo V, Jani S, et al. Catheter-based radiotherapy to inhibit restenosis after coronary restenting, N Engl J Med 1997; 336:1697–1703 10 Ip JH, Fuster V, Badimon L, et al. Syndromes of accelerated atherosclerosis: role of vascular injury and smooth muscle cell proliferation. J Am Coll Cardiol 1990; 15:1667–1687

1592

11 McCready RA, Hyde GL, Bivins BA, et al. Radiation-induced arterial injuries, Surgery 1983; 93:306 –312 12 Andros G, Schneider PA, Harris RW et al. Management of arterial occlusive disease following radiation therapy. Cardiovasc Surg 1996; 4:135–142 13 Kaluza GL, Raizner AE, Mazur W, et al. Long-term effects of intracoronary ␤-radiation in balloon- and stent-injured porcine coronary arteries. Circulation 2001; 103:2108 –2113 14 Waksman R. Late thrombosis after radiation: sitting on a time bomb. Circulation 1999; 100:780 –782 15 Smith RG, Potential advantages of treatment of transplanted saphenous vein aorto-coronary artery bypass grafts with beta irradiation to prevent graft occlusion. Artery 1997; 22:278 – 286 16 He GW, Ryan WH, Acuff TE, et al. Risk factors for operative mortality and sternal wound infection in bilateral internal mammary artery grafting. J Thorac Cardiovasc Surg 1994; 107:196 –202 17 Matsa M, Paz Y, Gurevitch J, et al. Bilateral skeletonized internal thoracic artery grafts in patients with diabetes mellitus, J Thorac Cardiovasc Surg 2001; 121:668 – 674 18 Suma H, Takanashi R. Arteriosclerosis of the gastroepiploic and internal thoracic arteries. Ann Thorac Surg 1990; 50:413– 416 19 Taggart DP, D’Amico R, Altman DG. Effect of arterial revascularization on survival: a systematic review of studies comparing bilateral and single internal mammary arteries. Lancet 2001; 358:870 – 875

Clinical Investigations