If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Is radiation-induced arteriopathy in long-term breast cancer survivors an underdiagnosed situation?: Critical and pragmatic review of available literature
Address: Oncologie-Radiothérapie-Radiopathologie, Groupe Hospitalier Universitaire, APHP site Saint-Louis- Université de Paris, 1, Ave Claude Vellefaux, 75010 Paris, France.
Breast node RT may result in RI arteriopathy after decades, with or without lymphedema.
•
Under-estimated axillary-subclavian RIA incidence by non-specific arm symptoms.
•
Ischemic arm pain for subclavian A, TIA for carotid A, or angina pain for coronary A.
•
CT or MR angiography can best reveal silent but threatening RI arteriopathy.
•
Short stenosis best treated by stent; long stenosis symptoms should be improved by aspirin.
Abstract
Purpose
Although considered exceptional, radiation-induced arteriopathy in long-term breast cancer survivors involves three main arterial domains in the irradiated volume, namely axillary-subclavian, coronary, and carotid. Stenosis of medium-large arteries is caused by “accelerated” atherosclerosis, particularly beyond 10 years after long-forgotten radiotherapy. The present review aims at summarizing what is known about arteriopathy, as well as the state of the art in terms of diagnosis and therapeutic management.
Diagnosis
Pauci-symptomatic over years, the usual clinical presentation of arteriopathy involves arm pain with coldness due to subacute or critical ischemia (arterial occlusion), wrongly attributed to an exclusive neurological disorder, and more rarely transient ischemic accident or angina. Evaluation of the supra-aortic trunks by computed tomography and/or magnetic resonance angiography visualizes artery lesions, while Doppler ultrasonography in expert hands assesses diagnosis and downstream functional impact. In severe cases, more invasive angiography directly visualizes long irregular arterial stenosis (full-field radiotherapy), allowing accurate prognosis and treatment.
Management
Requires early diagnosis to enable initiation of medical treatment that increases blood flow (aspirin) as soon as moderate stenosis is detected, combined with correction of vascular risk factors. In intermediate cases, these therapeutic measures are completed by revascularization strategies using transluminal angioplasty-stenting (wall thickness). Antifibrotic treatment is useful in advanced cases with combined radiation injuries.
Conclusion
In follow-up of long-term breast cancer survivors with node irradiation, myocardial infarction is treated even if radiotherapy is forgotten, while recognition and diagnosis of chronic arm ischemia due to subclavian artery stenosis needs to be improved for appropriate therapeutic management.
Radiotherapy (RT), combined with surgery and systemic treatments, is essential for cancer cure, allowing that half of treated patients are long-term survivors beyond 10 years of follow-up. However, RT may cause some long-term sequelae in the irradiated volume [
]. Diagnosis of these sequelae by non-oncological medical specialists as neurologists, cardiologists, or medical oncologists is dependent on recognition that long-past RT may be a causal factor, long after oncology follow-up has been stopped.
Some recent epidemiological studies have shown an excess of vascular deaths in women with BC who underwent adjuvant RT, while meta-analyses have established the founder role of RT in reducing local recurrences and improving cancer-free survival [
Early Breast Cancer Trialists’ Collaborative Group
Effect of radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year breast cancer death: meta-analysis of individual patient data for 10,801 women in 17 randomised trials.
]. After breast cancer (BC) lymph node irradiation, radiation-induced arteriopathy (RIA) involving axillary-subclavian, coronary, and carotid arteries, is little reported in the literature. Large vessel RIA is a poorly known and considered to be exceptional. However, its incidence could be under-estimated because of non-specific and delayed symptoms, combined with other confounding radiation-induced (RI) injuries, and falsely reassuring Doppler ultrasonography (US). By definition, RIA, within the irradiation field, concerns all arteries, the well-known small vessels supplying the irradiated volume itself, as well as the large vascular trunks leading to downstream injuries [
In the present review, new radiographic tools and clinical expertise are shown that could help the clinician in long term BC patient follow-up.
Diagnosis
In a patient with past oncology treatment, RI injuries are often diagnosed after a long medical course. Worse, diagnosis of RIA is not suspected, due to lack of symptom specificity, even if there are other combined RI sequelae in the same irradiated volume (Fig. 1).
Fig. 1Schematic representation of nodal fields in radiotherapy (axillary-supraclavicular and internal mammary chain) in women with right breast cancer with variable borders depending on the patient’s anatomy, protocols, and decade of treatment (equipment): axillary, subclavian, carotid-brachiocephalic trunk, possible coronary artery involvement.
The usual clinical presentation of RIA after BC lymph node irradiation is asymptomatic or involves mild symptoms that have evolved over years, mainly concerning the arm after supraclavicular RT. When arterial stenosis is progressive, chronic ischemia can be revealed by severe pain, especially after exercise (intermittent claudication) [
]. Palpable subcutaneous atrophy and fibrosis, downstream lymphedema, a raised area, or tattoo marks, may point to RI injuries. Asymptomatic linear radiopaque wall calcifications often indicate the arterial path in the irradiated volume.
Modrall and Sadjadi classified RIA according to time of onset: early (≤5 y) by wall thrombosis; intermediate (≤10 y) by wall fibrosis – occlusion in the absence of collaterality; and late (mean 26 y) by accelerated atherosclerosis with periarterial fibrosis [
]. Conversely, acute ischemia is revealed suddenly by distal gangrene (fingers, stroke) after arterial occlusion (thrombosis or distal embolus). Vascular assessment allows morphological diagnosis of the affected vessel, its downstream hemodynamic consequences and collaterality (quality of compensation). In fact, once arm pain is present, it should be approached by systematic computed tomography angiography (Fig. 2A-2B) and/or magnetic resonance angiography (Fig. 2C), which indirectly visualize axillary-subclavian and supra-aortic trunks, allowing noninvasive lesion check-up; followed by Doppler US in expert hands to describe the lesions and assess the downstream functional prognosis.
Fig. 2Radiation-induced axillary arteriopathy in long-term breast cancer survivors: A 73-year-old woman treated in 1981 by preoperative radiotherapy for left breast cancer consulted in 2015 for radiation-induced brachial plexopathy for 5 years, with moderate lymphedema, presternal calcinosis, and cold hands. (2A frontal −2B horizontal) Computed tomography angiography showed a prevertebral plaque, axillary 6 × 3 cm macrocalcification parallel to the chest wall (yellow arrow), leaning against a large periarterial fibrosis (yellow dotted line) sheathing a long irregular axillary artery stenosis (5 mm, red arrow) up to the pulmonary apex. Aspirin-pravastatin was begun. (2C) In 2017, successive skin bubbles appeared on the fingers and were wrongly attributed to cooking burns because of neurological anesthesia. Doppler US and magnetic resonance angiography confirmed axillary plaques (arrow) with moderate stenosis. In 2018, a second wave of bubbles was followed by arm erysipelas and neurological worsening. (2D) In 2019, a third wave of bubbles was diagnosed as subacute ischemia. Arteriography showed an axillary aneurysm upstream of a long stenosis including plaques, with thumb and forefinger embolus: percutaneous transluminal angioplasty was followed by placement of an 8–60 mm axillary stent.
Then, conventional angiography, which visualizes arteries directly, allows accurate invasive diagnosis and prognosis defining the severe arterial lesions and their extent (Fig. 2D). Axillary-subclavian RIA is characterized by a long (RT field size) irregular and moderately tight (50% over 6 cm) stenosis (Fig. 3A-Fig. 3B), but whose downstream flow is as problematic as with a very tight short (90% over 1 cm) stenosis (Fig. 3C).
Fig. 3Radiation-induced subclavian arteriopathy in long-term breast cancer survivors: (3A) A 70-year-old woman treated in 1982 by postoperative radiotherapy for left breast cancer consulted for radiation-induced brachial plexopathy since 2011. Because of arm coldness, Doppler US and computed tomography angiography were performed and the findings were “normal” in 2016. The patient's condition was improved by aspirin. Plexus magnetic resonance imaging with magnetic resonance angiography in 2017 confirmed a long (10 cm) 50% stenosis and the patient reported symptoms of neural sensitivity. (3B) A 72-year-old woman treated in 1981 by exclusive radiotherapy for right breast cancer consulted in 2012 for hand paresthesia. Moderate subclavian artery stenosis was treated by aspirin. In 2015, because of right radiation-induced brachial plexopathy, computed tomography angiography showed long (52 mm) axillary stenosis (yellow arrow: right diameter 3.2 mm, versus left 5.3 mm) inside thick perivascular fibrosis, combined with short prevertebral stenosis of the subclavian artery (65%, red arrow), with stable Doppler US findings. (3C) In 2018, cold hands and reduced radial pulse prompted magnetic resonance angiography, which showed both long postvertebral and short prevertebral arterial stenosis (70%, red arrow). (3D) Arteriography confirmed stable arterial stenosis (red arrow) and antifibrotic treatment combined with aspirin was implemented.
Radiological lesions include (i) pathognomonic long arterial stenosis, with short segmental occlusion in the absence of expected collaterals, with a very tight calcified stenosis characterized by wall and periarterial (external) fibrosis [
]; (ii) superficial wall thrombosis of a large plaque (internal); (iii) arterial aneurysm with possible distal embolisms, characterized by wall atrophy-necrosis.
Risk factors
The nature and severity of RIA depend on several interlinked factors [
] some related to older aspects of radiotherapy techniques, others concerning new options such as hypofractionation, any large RT volume, or a limited RT volume after an extensive surgery.
Treatment-dependent factors- Radiotherapy risk factors mainly involve technical parameters such as large total dose (≥50 Gy to vessel), high daily dose per fraction (≥2.5 Gy) in extensive volume [
] overlapping fields, re-irradiation in a given volume, heterogeneous high-dose distribution, hot spot high dose (field junctions, stereotaxy), low-energy RT machines (200 KV, cobalt) using a short source-to-skin distance by surface overdose, and patient movement favoring overlapping fields (position change or arm elevation between two fields during a given radiation exposure). In a recent prospective study, the risk of stenosis of the common carotid artery explored by Doppler US showed that intima-media thickness (IMT) after cervical RT is an independent vascular risk factor [
Surgery in the irradiated field enhances the risk of additional fibrosis (especially if there is hematoma, postoperative infection, or lymphatic stasis) or necrosis (delayed healing). Drugs such as angiogenesis inhibitors have an expected direct toxic vascular effect. Indirect effects such as Raynaud-like syndrome have also been described in testicular cancer survivors treated with bleomycin, etoposide, and platinum chemotherapy [
Patient-dependent factors- RI reactions are known to depend on the physiological condition of the patient (advanced age, obesity) and vascular comorbidity factors as arterial hypertension, uncontrolled diabetes, dyslipidemia, smoking, and vasculitis. Young age at the time of RT significantly increases the risk of RIA, as described in testicular cancer survivors in 8 patients with abdominal arterial stenosis versus 18 without stenosis, in a cohort of respectively 28 years old versus 38 at the time of RT [
All anatomical structures of the vascular system, from the aorta to capillaries, may be injured, as long as the vessel is included in the previously irradiated volume (Fig. 1). Moreover, the incidence of RIA between the 1960s and 1980s has been over-represented in studies whose technical modalities of imaging and RT (machine, positioning, dosimetry) were less precise and subject to more movement, and conversely under-estimated in aggressive cancers with short survival, in the absence of cohorts of long-term survivors. Each arterial site, axillary-subclavian, coronary, or carotid, concerns one of the 3 main types of arterial vascularization, respectively, the arm, heart, or head, with the related symptoms if there is stenosis. Patients may have one lesion but can develop others in the following months along the same trunk artery (axillary-subclavian) or in the other artery territory (coronary).
Radiation-induced damage to the axillary-subclavian arteries
RIA of the supra-aortic trunks is observed in the follow-up of patients after RT of the axillary-subclavian lymph nodes mainly for BC, and less so after cervical-subclavian RT for head and neck cancer (HNC), lymphoma, sarcoma, thyroid, or apical cancer.
Although, axillary-subclavian RIA is not foremost in the doctor’s mind when a patient is suffering from arm pain, axillary or subclavian artery stenosis might translate into several clinical pictures:
(a)
asymptomatic, during a check-up for RI brachial plexopathy or lymphedema;
(b)
stress ischemia, upper limb exercise-induced claudication with muscle pain, cramping, and fatigue during daily movements, worse during arm elevation [
Radiotherapy-related axillary artery occlusive disease: percutaneous transluminal angioplasty and stenting. Two case reports and review of the literature.
subacute ischemia, the main mode, preceding an acute event by a few months, including digital pain at rest, coldness, loss of color revived by rest, periungual and pulp wound, and a numb and clumsy hand; clinically relevant because of its insidiousness (few symptoms for a long time).
(d)
critical ischemia by arterial occlusion, with sudden pain at rest, pallor or cyanosis, coldness and acral paresis because of embolic lesions of the fingers due to an ulcerated plaque [
]. Bilateral humeral blood pressure and palpation of radial pulses may reveal an asymmetry in subclavian artery stenosis. Doppler US is often falsely reassuring because without expertise the subclavian area is hard to access.
Axillary-subclavian RIA is a misleading clinical expression related to the crossroads of the supra-aortic trunks: presence of arterial replacement pathways (subclavian, vertebral, carotid, internal mammary) and modest muscle requirements in blood flow. Thus, a hemodynamically significant chronic subclavian stenosis of more than 50% will have circulatory expression in the arm, rather than retrograde circulation in the vertebral artery (subclavian steal syndrome) [
A randomized, placebo-controlled, clinical trial combining of pentoxifylline- tocpherol and clodronate in the treatment of radiation-induced plexopathy.
] 28 out of 37 patients had these nonspecific vascular symptoms related to tight RIA stenosis over 1 cm and/or stenosis over 4–8 cm. The time lapse between RT and RIA was 28 ± 9 years, and RIA was clearly symptomatic within the decade following the occurrence of RI brachial plexopathy, with tight stenosis > 75% -occlusion in 11 cases and long moderate stenosis, mainly located at the scaleno-vertebral triangle in 17 cases [
Delanian S, Klein I, Dadon M, et al. Axillo-subclavian entrapment in radiation plexitis revealed throughout a randomized trial. ESTRO 37 meeting, Barcelone, Spain. 20- 24/04/18 (EP1284). Radiother Oncol 2018; 127: S706
]. Thus, if paresthesia of the hand or with the cervico-brachial path is the expression of sensory neurological impairment, the pain reported by the patient in the hand, forearm, or shoulder is instead suggestive of associated chronic ischemia and should be assessed by computed tomography angiography and Doppler US during oncologic follow-up. Note that significant arterial stenosis is observed in 60% of patients (28 of 47) combined with brachial plexopathy.
In contrast, in critical ischemia caused by acute occlusion in the upper limbs, the first hypothesis is related to heart disease embolism. However, in irradiated patients, the embolism is likely due to an arterial ulcerated plaque or a proximal aneurysm (Fig. 2D). The embolism is humeral (60% at the bend of the elbow) or acral, while the severity of ischemia is related to the volume of the embolus and the potential for collaterality. There may be complete arterial occlusion in a chronic tight RIA stenosis, with the occurrence of partial or total abrupt upper limb paralysis. Acute arterial stenosis in a recent traumatic context may also reveal an artery transfixed by a displaced clavicle fracture in osteoradionecrosis [
RIA in the axillary-subclavian volume is poorly documented in the literature and its incidence in BC survivors, said to be exceptional, is not known. Thirty-nine publications between 1973 and 2015 comprise only 127 patients in case reports of highly advanced RIA and two cohort studies [
]. RIA was noted in 103 (81%) cases after BC (Table 1), 15 after lymphoma (mantle RT), 6 after HNC, 2 after sarcoma, and 1 case after lung cancer. Among these 103 BC patients, most (98) were suffering from ischemia due to arterial stenosis-occlusion: subclavian arteries in 83 cases, combined with brachiocephalic trunk-carotid stenosis in 6 [6, 11; 14–19; 24–38], and axillary arteries in 15 [
Radiotherapy-related axillary artery occlusive disease: percutaneous transluminal angioplasty and stenting. Two case reports and review of the literature.
A case of axillary arterial bleeding caused by radiation-induced chest wall ulcer after radiotherapy for carcinoma of the breast: Extraanatomic bypass grafting for upper limb salvage.
]. Their mean age was 65.3 ± 12.7 years, with a mean time from RT to RIA of 16.1 ± 10.8 years (Table 1). Patients presented with ischemic pain as arm claudication in 3 cases, upper extremity subacute ischemia in 66, and critical ischemia in 28. Two-thirds of cases showed RI brachial plexopathy (15 cases, ± acute paresis), or osteoradionecrosis, skin ulcer, arm lymphedema. Treatment was vascular surgery in 56 cases (bypass, some amputations) until 2000, and subsequently more conservative using percutaneous transluminal angioplasty-stenting in 10 cases [
] and medical treatment in 8. Only five of the 103 cases involved hemorrhage after skin ulcer, epistaxis, hemoptysis, or pseudoaneurysm by arterial rupture (subclavian artery in 1, axillary artery in 2, brachiocephalic trunk in 2), and were treated by bypass surgery in 1 case [
A case of axillary arterial bleeding caused by radiation-induced chest wall ulcer after radiotherapy for carcinoma of the breast: Extraanatomic bypass grafting for upper limb salvage.
Table 1Radiation-induced arteriopathy in long-term breast cancer survivors: axillary-subclavian incidence based on 103 case reports over 42 years: 84 cases of subclavian (±6 brachiocephalic trunk) artery disease, 17 of axillary artery disease, and 2 of brachiocephalic trunk artery disease.
A case of axillary arterial bleeding caused by radiation-induced chest wall ulcer after radiotherapy for carcinoma of the breast: Extraanatomic bypass grafting for upper limb salvage.
Radiotherapy-related axillary artery occlusive disease: percutaneous transluminal angioplasty and stenting. Two case reports and review of the literature.
Finally, assessment using modern vascular tools and specific care of the homolateral arm in BC survivors, 15–30 years after node RT, suggests that axillary-subclavian RIA stenosis can be significant, possibly combined with osteonecrosis or plexopathy. It should be explored by computed tomography angiography to discover whether medical or radiological treatment is needed.
Radiation-induced damage to carotid arteries
Carotid RIA (Table 2) is really exceptional in follow-up to unilateral node RT for BC, and more common and well-described after bilateral cervical-subclavian RT for HNC, lymphoma or thyroid cancer [
Effect of nodal irradiation and fraction size on cardiac cerebrovascular mortality in women with breast cancer with local and locoregional radiotherapy.
Table 2Radiation-induced arteriopathy in long-term breast cancer survivors: incidence of carotid artery disease based on epidemiological studies between 2005 and 2011.
Study (year) [ref]
Nb CVA-TIA cases (Nb RT nodes) Total of cancer cases
Effect of nodal irradiation and fraction size on cardiac cerebrovascular mortality in women with breast cancer with local and locoregional radiotherapy.
The interval from RT to symptomatic extra-cranial carotid RIA is variable. The clinical presentation is mostly delayed cerebrovascular accidents: transient ischemic accident or stroke. Morphological diagnosis is approached by magnetic resonance angiography focused on the supra-aortic vessels, while Doppler US should assess carotid RIA stenosis.
In fact, the homolateral carotid artery is very little affected by BC irradiation. First, because part of the carotid artery is positioned at the edge of the supraclavicular RT volume used to treat BC patients, the carotid artery generally receives less than the mean 50 Gy prescribed dose. Second, the full length of the homolateral carotid artery is not included in the supraclavicular RT volume. Moreover, in literature reports, the supraclavicular beam is usually combined with the mammary chain beam, which includes the brachiocephalic trunk and common carotid. And the homolateral carotid artery seems to be the site of rare cases of carotid RIA, though this thoracic part of the artery is difficult to assess by Doppler US and magnetic resonance angiography (Fig. 4).
Fig. 4Radiation-induced carotid arterial disease in long-term breast cancer survivors: (4A) A 68-year-old woman treated in 1988 by postoperative radiotherapy for right breast cancer consulted in 2015 for recent radiation-induced brachial plexopathy and arm lymphedema. Severe hand coldness prompted computed tomography angiography in 2018 which revealed subclavian artery stenosis, confirmed by Doppler US, with very significant hemodynamic carotid artery stenosis (5 m.sec-1). Magnetic resonance angiography confirmed right common carotid stenosis (90%, arrow). (4B) Arteriography (arrow) showed a short occlusive stenosis with downstream consequences. Percutaneous transluminal angioplasty and placement of a 7 × 18 mm carotid artery stent were performed without complications. (4C) An 83-year-old woman treated in 1976 by exclusive radiotherapy for left breast cancer consulted in 2017 for threatening presternal subcutaneous calcinosis. Vertigo prompted computed tomography angiography, which revealed several tight arterial stenoses: L-subclavian (70%), L-vertebral (90%, followed by stenting), and calcified occlusive right brachiocephalic trunk (arrow) treated medically. (4D) A 79-year-old woman with postoperative radiotherapy for left breast cancer in 1982 was treated for chronic skin wounds and sternal osteoradionecrosis present since 2013. Vertigo in 2018 prompted vascular explorations, which revealed: postvertebral left subclavian and left vertebral (right compensation) artery occlusions, left common carotid artery (70%) and right brachiocephalic trunk stenosis (80%). Placement of a 7 × 18 mm stent in the brachiocephalic trunk (arrow, angiography) led to clinical improvement without stroke.
]. However, carotid RIA after regional BC irradiation of the cervicothoracic lymph nodes (supraclavicular – mammary chain) is poorly documented and its incidence in BC survivors is debated.
After breast cancer, carotid RIA was reported in 1978 in 3 cases out of 9 women aged 52 to 70, 2–24 years after supraclavicular RT (kilovoltage irradiation; mean dose 5500 rads). Amaurosis or stroke revealed carotid stenosis at the bifurcation in 2 patients and common carotid-subclavian-brachiocephalic trunk stenosis in one patient; the patients were treated by endarterectomy [
]. A 42-year-old woman who experienced acute pain in the right arm and left hemiparesis, 5 years after BC, was found to have occlusion of the right internal carotid and subclavian arteries and was treated by thrombectomy [
]. However, these differences in the incidence of carotid RIA after BC are technical rather than epidemiological: the RT of BC lymph nodes does not strictly cover the same volume, depending on the country, institution, and time period concerned. In BC cervical RT, technical variants of the direct beam of the supraclavicular area used an oblique field and tracheal protection with rotation of the patient’s head, which was not very reproducible [Fig. 1]: the common carotid artery was partially irradiated along its path and diameter. The first centimeters of the common carotid, in its intrathoracic course (left directly from the aortic arch and right from the brachiocephalic trunk) are only affected by the internal mammary chain (IMC) beam. Some angiographic characteristics have been described for carotid RIA stenosis [
]. In a recent epidemiological study, no significant difference was shown between 173 BC survivors and 346 matched cancer-free women in carotid IMT, while survivors had lower cerebral blood flow. In this study, carotid artery was partly in the IMC beam (2/3), while large carotid part related with supraclavicular RT beam (1/3) was associated with higher IMT [
More generally, after any cervical RT, there is a well-described excess risk of cerebrovascular accident because of carotid artery occlusion-ischemia, suggesting a low risk after BC carotid irradiation. Seven of 9 studies (2002–09) reported an increased incidence of stroke, as attested by a relative risk (RR) of 5.6 after HNC [
]. The incidence of 10-year cerebrovascular events in 6862 cases of HNC was 34% after RT alone (mean 60 Gy) versus 25% for surgery alone or combined with RT (p < 0.01) [
]. In 1926 Hodgkin lymphoma survivors irradiated in childhood between 1970 and 1986, 24 had a stroke: RR 5.6 (2.6–12.2) after mantle RT (mean 40 Gy). After adjuvant treatment of BC, radiation (mean 50 Gy) to the supraclavicular nodes and internal mammary chain increases the risk of stroke. The first study in 2005 [
]. In a meta-analysis of these studies, cervical RT at least doubled the RR of transient ischemic accident or stroke, with the exception of BC treated by minimal radiation exposure [
After any cervical RT, the risk of asymptomatic carotid RIA stenosis was initially described by angiography with a prevalence of 17–25%. It was documented further by Doppler US [
]: in 9 of 10 studies, a 16 to 55% prevalence of carotid RIA stenosis mostly at the carotid bifurcation artery. In asymptomatic HNC survivors after hemi-neck RT, there is an increased risk of carotid RIA stenosis (8 RI/44 versus 3 non-irradiated/44) [
] while Hodgkin lymphoma follow-up by cervical Doppler US showed a risk of RI CAS of 19% versus 2% in controls. A prospective analysis in BC, after supraclavicular RT, found no excess risk of carotid RIA stenosis in 46 patients at 15 years [
Prospective analysis of carotid artery flow in breast cancer patients treated with supraclavicular irradiation 8 or more years previously: no increase in ipsilateral carotid stenosis after radiation noted.
]. After cervical RT, the carotid intima-media thickness, assessed by Doppler US, is interesting because the intermediate phenotype for early atherosclerosis correlates well with histology [
]. After hemi-cervical RT, in 42 HNC patients, a 35% (0.30 mm) increase in carotid IMT was measured at 100 months with RT versus non-RT contralateral control [
]. An ongoing prospective study evaluating carotid IMT by magnetic resonance imaging has shown that silent brain infarction occurred 5 times more frequently than stroke-transient ischemic accident [
Long term cerebral and vascular complications after irradiation of the neck in head and neck cancer patients: a prospective cohort study: study rationale and protocol.
Finally, the risk of carotid RIA after BC lymph node irradiation is exceptional, but does exist for a limited part of the artery inside the RT volume, mainly thoracic around the brachiocephalic trunk. Diagnosis is difficult, without the help of Doppler US. Carotid RIA should be considered during exploration of the supra-aortic trunks for arm ischemia or transient ischemic accident.
Radiation-induced damage to coronary arteries
Initially considered radiation-resistant and possibly sensitive to 30 Gy radiation after lymphoma, the heart clearly exhibits late RI toxicity, which was mostly described after the year 2000 [
]. Late coronary RIA is to date well documented in the literature in two situations: in earlier studies after mediastinal “large volume-low dose” RT for lymphoma-seminoma, and more recently “small volume-high dose” after thoracic parietal RT for BC [
]. While the risk of death from ischemic heart disease has substantially decreased over time with improvement in RT techniques, patient follow-up over several decades has always shown an increased risk of heart disease [
The clinical presentation of coronary RIA is usually asymptomatic or non specific angina.
Numerous case reports of coronary RIA were published between 1958 and 2013: 44 publications including 66 patients (mean age 40 ± 14 y) irradiated a mean 6.4 ± 6 y before, for lymphoma in 70% of cases, and free of vascular risk factors. Coronary angiography provided evidence of long and proximal coronary stenosis, sometimes misinterpreted because of parietal calcifications, which interfere with the visualization of vascular light (over- and under-estimates) [
Coronary artery disease after radiation therapy for Hodgkin's lymphoma: coronary CT angiography findings and calcium scores in nine asymptomatic patients.
]. In lymphomas, the mantle RT volume concerned the right and left anterior coronary arteries. Beyond case reports describing the first cases of coronary RIA after high dose-volume RT, low dose-volume RT has since been shown to reduce this risk: from mantle RT at 40–44 Gy in the 1960s (RR mortality 6.3), then mantle RT at 30–40 Gy with combined chemotherapy of the 1980 s (RR 1.97), to involved-field RT at 20 Gy from the 1990s [
]. In a prospective French study with evaluation by computed tomography coronary angiography, out of 179 patients followed up for 10 years, 46 (26%) had coronary RIA: 15% at 5 years, 34% at 10 years, with 12 cases of severe non-ostial stenosis requiring a local procedure [
Breast irradiation and coronary RIA concerned left breast RT in the thoracic parietal volume and mammary chain RT in the retrosternal volume, recently documented in BC survivors [
Mortality from myocardial infarction after adjuvant radiotherapy for breast cancer in the surveillance, epidemiology, and end-results cancer registries.
]. Published cases are old and rare (Table 3): 14 women, between 1958 and 2013, mean age 50 ± 5 y, irradiated 4.7 ± 3.3 y before, mostly revealed by angina [37; 83–90]. Whereas RT planning has improved over past decades by reducing cardiac dose exposure, problems persist [
]. Epidemiological studies of BC patients treated in the 1980 s and 1990 s demonstrated a significant increase in the incidence of RI heart disease in left-sided versus right-sided BC. RT of the left breast (tangential wall beams) and/or IMC (direct beam) exposes adjacent coronary arteries (Fig. 5). Quantification in 50 patients of contemporary RT doses received by the heart and coronary arteries found a mean 7.6 Gy (left anterior descending [LAD] coronary artery) and 2.3 Gy (heart) for left BC versus 1.2 to 2 Gy for right BC, and part of the heart volume > 20 Gy was observed in half of left BC patients [
]. Over all 398 regimens reported in 149 studies from 28 countries, estimations of mean heart dose in left BC were 5.4 Gy (range 0.1–28.6), 4.2 Gy without IMC versus 8 Gy with IMC, and in right BC 3.3 Gy, with wide variation between studies, even for apparently similar regimens [
Table 3Radiation-induced arteriopathy in long-term breast cancer survivors: incidence of coronary artery disease based on 14 case reports over 55 years.
Fig. 5Radiation-induced coronary arteriopathy in long-term breast cancer survivors: (5A) Computed tomography scan dosimetry of tangential radiotherapy fields for left breast cancer, with variable borders depending on the patient’s anatomy and the protocol: left anterior descending branch and variable part of left ventricle were included in the radiotherapy breast volume, while the right coronary artery was affected by an internal mammary chain field. (5B) A 67-year-old woman treated in 1989 by postoperative radiotherapy for left breast cancer consulted in 2013 for left radiation-induced brachial plexopathy present for 5 years. A first vascular exploration because of severe hand pain and collateral venous circulation was negative. In 2018, after stopping effective antifibrotic treatment, she suffered from severe hand pain. Magnetic resonance angiography showed minimal left axillary artery stenosis with calcified plaque, and arteriography showed left palm arterial arcade blockage by an embolus: placement of a 7 × 30 mm stent led to delayed reduction in pain. In parallel, pre-anesthesia electrocardiogram abnormalities led to coronary angiography, which showed severe stenosis (70–90%) of the left anterior descending coronary artery. (5C) Double stenting (2.5 × 20 mm then 3 × 12 mm) of the left anterior descending coronary artery was performed, without any complications.
A Swedish study evaluated by coronary angiography the link between coronary stenosis and BC radiotherapy techniques in 199 patients between 1990 and 2004. RT “hot spot” volumes were very well defined: for the proximal right coronary artery, after left mammary chain RT, and for the middle-distal LAD coronary artery, and for the distal diagonal coronary artery after left breast tangential RT. The odds ratio of tight coronary stenosis or grade 4–5/5 occlusion was 7.22 for the left breast versus right breast, in the “hot spot” zone [
]. The incidence and distribution of coronary RIA was assessed by cardiac stress testing and/or catheterization: right and left BC had the same estimated 7% 10-year risk, while at 12 y (2–24) there was a statistically significantly higher prevalence of abnormalities for the left BC (27/46) versus right BC (3/36), with 70% in the LAD coronary artery territory [
]. A retrospective study of 230,000 patients showed an RR of infarction of 5.28 at 10–15 years in women < 60 y after RT for left BC versus right BC. However, after 1980 the risk of coronary RIA after BC greatly decreased after systematic IMC irradiation was stopped [
]. A recent meta-analysis of 289,109 patients/13 studies established an RR of coronary RIA death of 1.23 for left BC versus right BC at 15 years. Finally, a case-control study of major coronary events (infarction, coronary revascularization, and cardiac ischemia death) in 2168 women between 1958 and 2001 determined coronary RIA risk: mean whole-heart dose of 4.9 Gy (0.03–27.72), linear cardiac risk without threshold of 7.4% per Gy (2.9–14.5); the increased risk started within the first 5 years after RT and continued over three decades [
Finally, there is a small life-long risk of coronary RIA after left BC, even after 10 Gy or less, which progressively decreases over decades, mainly in relation to mammary chain (LAD coronary artery) or tangential field (right coronary artery) RT.
Pathophysiology
RIA damage, first described in 1899 by Gassman, was considered as the first early stage of RI complications. Large artery stenosis was reported in 1959 [
]. It was subsequently realized that, besides intrinsic vascular damage, there is also extrinsic perivascular late injury related to RI fibrosis of the surrounding tissues. RIA in descending order of frequency, affected the following blood vessels with different clinical consequences: sinusoid capillaries (richest in endothelial cells); arterioles (simplified wall, ≤0.1 mm); medium muscular arteries (5 mm); large elastic arteries (>5 mm); small muscular arteries (2 mm), and large veins [
Capillary and arteriolar RIA has been well described involving direct injury to tissue rich in stroma. The extreme radiosensitivity of endothelial cells underlies acute RI effects, with early inflammation, modification of capillary permeability, thrombosis, focal loss of capillaries, followed by ischemic necrosis [
]. Later, this microcirculatory impairment depletes collateral blood vessels, resulting in telangiectasias, while irradiation of small arteries leads to fibrosis, lipid accumulation in macrophages, and possible luminal occlusion.
Large vessel late RIA is different, and indirect, reflecting the clinical arterial distribution: an anastomotic network in the case of occlusion of a branch (limbs), or a terminal mode with independent branches for parenchyma, but whose occlusion causes necrosis (brain, kidney) [
]. The vascular wall of medium-size arteries (8–12 mm) comprises three layers: intima (endothelium-basal membrane), media (smooth muscle cells – elastic matrix) dependent on circulating neuromediators, and adventitia (dense fibroblastic matrix) with the nutrient vasa vasorum of two-thirds of the external wall and sympathetic and parasympathetic parietal innervation.
RIA Pathogenesis – Although solidly embedded in a thick matrix, large RIA may present myointimal proliferation with lipid deposition, and wall thrombosis occlusion or rupture, associated with extensive periadventitial RI fibrosis. Experimental results obtained after irradiation of canine femoral arteries allowed characterization of the pathogenic process showing endothelial flaking and the appearance of foam cells (empty) and acute intimal impairment, followed by media fragmentation, extensive fibrosis of the three parietal layers, and then periarterial spreading [
]. RI accelerated atherosclerosis was initially described in rabbits after a hyperlipidic diet, and then in genetically modified mice with additional vascular risk factors, as in apolipoprotein E knockout mice submitted to cervical RT and increased carotid plaque [
]. However, RIA differs quantitatively and qualitatively from age-related atherosclerosis in terms of increased proteoglycan content and inflammatory cells in irradiated intima [
There are four complementary approaches to the management of established RIA depending on its severity: control of vascular risk factors, blood flow “fluidity”, reduce vascular wall thickness, and reduce perivascular RI fibrosis. For prevention, to reduce coronary RIA following future breast RT, it is recommended to exclude the heart volume of left-sided RT, and to exclude the root of the aortic arch where the right and left coronary arteries originate when irradiation of the IMC volume is indicated.
Control of vascular risk factors- The management of RIA involves in all cases the control of vascular risk factors: smoking cessation, control of high blood pressure, diabetes, and dyslipidemia, lifestyle changes (sedentariness), and reduction of overweight. Atherosclerosis prevention by the use of statins to lower cholesterol is very interesting in our patients, as pravastatin reduces RI lesions by decreasing chemotactic and pro-inflammatory substances. A French trial in 135 patients treated for carotid RIA stenosis showed that statin use was a predictor of fewer neurological events compared to no treatment [
Pravastatin reverses established radiation-induced cutaneous and subcutaneous fibrosis in patients with head and neck cancer: results of the biology-driven phase 2 clinical trial Pravacur.
]. For primary prevention in high-risk asymptomatic tight stenosis, long-term use of aspirin and/or antiplatelet agents as clopidogrel is recommended and useful in most of our patients after vascular assessment. After revascularization, the compromise between thrombotic and hemorrhagic risks od drugs suggests the use of aspirin and clopidogrel combination therapy for one month after insertion of a naked stent, for 6 months after insertion of an active stent, and for 12 months after bypass surgery, followed usually by aspirin monotherapy. Anticoagulant treatment targeting nerve chronic ischemia has also been debated: a patient with RI brachial plexopathy was helped by 3-month antivitamin K treatment, with transient regression of upper limb weakness and disappearance of conduction blocks on electromyography [
]. However, anticoagulant (heparin) use is reserved for acute events in the arm, head, or heart. The anticoagulants showed better efficacy in inhibiting platelet cohesion and fibrine deposition on procoagulant surfaces, than on the collagen surface of arterial surface wall [
Reduction of vascular wall thickness- Revascularization is the treatment of choice for advanced arterial stenosis, followed by prolonged double antiplatelet treatment. Its indication is based on the severity and symptomatic character of the stenosis, and on the possible downstream consequences after arterial occlusion, such as brain risk for a carotid artery, amputation or arm paralysis risk for a subclavian artery. The choice of revascularization technique is related to the viability of the irradiated tissues, in relation to deficient arterial collaterality and healing tissue defects.
Endoluminal interventional radiology- Today, percutaneous transluminal angioplasty, an endovascular treatment, is the best way to treat RIA (1st in 1982), as it avoids surgical risk related to access through irradiated tissue [
] (Fig. 3D; Fig. 4D; Fig. 5C). After dilatation, the arterial stent reduces the risk of immediate arterial dissection-rupture. This “naked stent” also prevents the risk of immediate arterial restenosis after dilatation, as it covers the RIA both upstream and downstream. The use of an “active stent”, covered with a cytostatic drug, reduces the risk of restenosis, thrombosis, and local inflammation. Restenosis is more frequent and occurs faster in the case of RIA versus atherosclerosis.
For the carotid artery, asymptomatic RIA stenosis > 80% or symptomatic RIA stenosis > 50% is treated by stenting or surgery if the cervical subcutaneous tissues allow. Symptomatic stenosis (transient ischemic attack-stroke within 6 months) > 70% requires urgent revascularization. A meta-analysis of 27 studies analyzed the treatment of 533 cases of carotid RIA stenosis, 361 by stenting (2000–2011) and 172 by endarterectomy (1978–2011). Non-restenosis (50%) was significantly higher after stenting versus endarterectomy, without the risk of postoperative complications [
]. The overall rate of restenosis (more than 50%) after carotid RIA stenting was 5.4% person-years (95% CI 4.3–6.6), but large differences were observed among studies at 30 months: no restenosis was reported in two small studies (5 and 7 patients), 17–21% in two larger studies (16 patients), and 12% in the largest one (18/135) [
For the coronary RIA, tight stenosis is treated by stenting. While the indication for aortocoronary bypass surgery has decreased because of postoperative RI complications, the risk of coronary RIA thrombosis after stenting is twice that for control coronary patients or for dissection [
]. However, it seems that vascular treatment has a role in the control of pain in RI brachial plexopathy, even without clear ischemic symptoms (personal observation).
Vascular surgery- For a long time, bypass grafting was the surgical technique of choice for old RI scleroatrophic lesions, using non-irradiated healthy wall anastomoses, as dissection of periarterial fibrous tissue is difficult and dangerous [
Subclavian-axillary graft plus graft-carotid interposition in symptomatic radiation-induced occlusion of bilateral subclavian and common carotid arteries.
]. Venous or arterial bypass surgery is preferable for long and irregular RIA lesions and is performed using venous or prosthetic grafts, which are robust to postoperative compression, but open to the risk of infection [
Reduction of perivascular fibrosis- The fibrosis appearance in RIA is always present and is recognized after a vascular emergency. We think that a corticosteroid treatment, which has long been used to limit the inflammatory phenomena associated with fibrosis, is useful in acute RIA, in combination with anticoagulant, for a mean one-month period [personal observation]. In a more chronic phase, and because of its local perivascular compression, RI fibrosis is a therapeutic target (Fig. 2A). The PENTO synergistic combination (pentoxifylline-tocopherol) is useful because it reduces RI fibrosis after 6 to 18 months of treatment [
]. Clodronate, a bisphosphonate known for its inhibition of osteoclastic destruction and its chronic anti-inflammatory activity, has been shown to amplify the PENTO antifibrotic effect (in a PENTOCLO combination) in refractory osteoradionecrosis or RI plexopathy [
A randomized, placebo-controlled, clinical trial combining of pentoxifylline- tocpherol and clodronate in the treatment of radiation-induced plexopathy.
Complete restoration of refractory mandibular osteoradionecrosis by prolonged treatment with a pentoxifylline-tocopherol-clodronate combination (PENTOCLO): a phase II trial.
] and is useful in severe and complex RI injuries.
RIA is a heterogeneous entity that is increasingly reported thanks to the latest medical imaging. Long-term cancer survivors are likely to develop late complications of RIA, even after the recent development of focused radiation therapy. This review tried to provide tools (medical expertise, imaging, literature) to help diagnosis and therapeutic management of RIA in order to enhance the BC survivors’ quality of life. Whereas myocardial infarction is well managed, even if long-past irradiation is forgotten, stenosis of the thoracic carotid and subclavian arteries is currently not diagnosed and treated enough. Vascular therapeutics and indications suggest options that will be more effective when introduced early, as aspirin, pravastatin, and arterial stenting.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
We thank Dr S Awad for expertise in radiation-induced vascular disease, Prof PF Pradat for neurological expertise, J-L Lefaix for help in understanding the pathophysiology of arteriopathy, and Prof M-C Vozenin for her clinical and biological expertise.
Early Breast Cancer Trialists’ Collaborative Group
Effect of radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year breast cancer death: meta-analysis of individual patient data for 10,801 women in 17 randomised trials.
Radiotherapy-related axillary artery occlusive disease: percutaneous transluminal angioplasty and stenting. Two case reports and review of the literature.
A randomized, placebo-controlled, clinical trial combining of pentoxifylline- tocpherol and clodronate in the treatment of radiation-induced plexopathy.
Delanian S, Klein I, Dadon M, et al. Axillo-subclavian entrapment in radiation plexitis revealed throughout a randomized trial. ESTRO 37 meeting, Barcelone, Spain. 20- 24/04/18 (EP1284). Radiother Oncol 2018; 127: S706
A case of axillary arterial bleeding caused by radiation-induced chest wall ulcer after radiotherapy for carcinoma of the breast: Extraanatomic bypass grafting for upper limb salvage.
Effect of nodal irradiation and fraction size on cardiac cerebrovascular mortality in women with breast cancer with local and locoregional radiotherapy.
Prospective analysis of carotid artery flow in breast cancer patients treated with supraclavicular irradiation 8 or more years previously: no increase in ipsilateral carotid stenosis after radiation noted.
Long term cerebral and vascular complications after irradiation of the neck in head and neck cancer patients: a prospective cohort study: study rationale and protocol.
Coronary artery disease after radiation therapy for Hodgkin's lymphoma: coronary CT angiography findings and calcium scores in nine asymptomatic patients.
Mortality from myocardial infarction after adjuvant radiotherapy for breast cancer in the surveillance, epidemiology, and end-results cancer registries.
Pravastatin reverses established radiation-induced cutaneous and subcutaneous fibrosis in patients with head and neck cancer: results of the biology-driven phase 2 clinical trial Pravacur.
Subclavian-axillary graft plus graft-carotid interposition in symptomatic radiation-induced occlusion of bilateral subclavian and common carotid arteries.
Complete restoration of refractory mandibular osteoradionecrosis by prolonged treatment with a pentoxifylline-tocopherol-clodronate combination (PENTOCLO): a phase II trial.
00009-8&id=gr1.jpg){kind=link}
00009-8&id=gr2.jpg){kind=link}
00009-8&id=gr3.jpg){kind=link}
00009-8&id=gr4.jpg){kind=link}
00009-8&id=gr5.jpg){kind=link}