Accelerated atherosclerosis and carotid stenosis are well-established risks occurring after high radiation doses that are used to treat cancers of the head and neck. Noncoronary vascular disease has been observed and may relate to more moderate dose irradiation.
A search of patients treated for Hodgkin disease, non-Hodgkin lymphoma, or seminoma was performed to identify cases with noncoronary vascular complications after irradiation. These three groups were chosen because of the use of intermediate dose radiation and prevalence of long-term survivors. Individual patient records were reviewed to document the type and presentation of the stenosis and the clinical factors that may have contributed to this risk.
Twenty-one patients were identified who developed disease in noncoronary arteries after treatment. The median time from irradiation to diagnosis of vascular stenosis was 15 years. Antecedent risk factors for vascular disease were prevalent. Five patients had disease identified by auscultation of bruits before an adverse clinical event occurred. Five patients died from complications related to their vascular disease, which included three deaths after stroke and two after small bowel infarction.
As cancer treatment has continued to evolve there is an increasing number of long-term survivors after radiotherapy. As a result, the diagnosis and management of long-term radiation-induced complications have become more important. A rare yet potentially serious complication of radiation is large vessel injury. In 1959, Thomas and Forbus1 reported the first known case of irradiation injury to a large vessel in a young man with non-Hodgkin lymphoma who was found to have necrosis of the aortic wall at autopsy.
Radiation injury to large arteries can result in vessel rupture, stenosis, or occlusion. Rupture may occur early secondary to arterial wall necrosis.2 Fajardo and Lee3 reported on 11 patients who developed rupture of major vessels after irradiation. They concluded that surgical complications were the most important cause of rupture and that only 2 of the 11 cases appeared to have been caused mainly by radiation.
The small capillary and sinusoid vessels are probably the most sensitive to radiation because endothelial cells are quite radiosensitive and make up most of the vascular wall. Although endothelial cells of large arteries also undergo extensive radiation injury, their function is not immediately affected because they have a strong vascular wall and an ample lumen.4 Nonetheless, long-term radiation damage to large arteries can produce dramatic effects such as stroke, myocardial infarction, or mesenteric infarction.
The development of stenosis or occlusion of large arteries has been attributed to the development of fibrous intimal plaques.2, 5 Accelerated atherosclerosis and carotid stenosis have been documented after radiation doses greater than 50 Gy that are used to treat epithelial carcinomas of the head and neck.6–9 However, there is evidence for an increased risk of coronary artery disease from more moderate dose irradiation in survivors of Hodgkin disease.10, 11 Reports of noncoronary atherosclerotic vascular disease after moderate dose irradiation (30–50 Gy) have been sporadic and the complication is not well understood.11–13
Some radiotherapists, surgeons, and medical oncologists are aware of this long-term possibility, but many physicians may not have had personal experience with these lesions. Therefore, it was felt that these rare noncoronary vascular lesions developing after radiation should be described in detail. Documentation of these late complications is important for surveillance and intervention, but ultimately for the prevention of late vascular effects.
MATERIALS AND METHODS
A computerized search of 7165 followup records for patients treated for Hodgkin disease (3483), non-Hodgkin lymphoma (3422), or seminoma (260) treated at Stanford University Medical Center between 1959 and 1991 was performed to identify cases with noncoronary atherosclerotic disease after irradiation. These three groups were chosen because of the use of intermediate dose irradiation and the high prevalence of long-term survivors among this patient population. Intermediate dose is defined as 30–50 Gy prescribed dose. Noncoronary atherosclerotic disease was defined as 40% or more stenosis of a large noncoronary artery. The vascular complication must be within the radiation field for the patient to be included in the study. A total of 21 patients formed the cohort of this study. Because follow-up, evaluation, and complication encoding was likely to vary, we could not estimate the true prevalence or incidence of these events.
The median follow-up was 19.8 years (range, 1.3–44.3 yrs). Follow-up started on the day of the last radiation treatment and concluded on the date of death or last follow-up visit. The median time to a vascular event was 15 years (range, 1.3–40 yrs). The time to event was defined as the period from the date of the last radiation treatment until the date of the imaging study that documented the large vessel stenosis or occlusion.
Information was obtained retrospectively through radiation therapy and hospital charts and physician records. The clinical patient characteristics examined included: 1) age at treatment; 2) age at diagnosis of vascular disease; 3) presence or absence of the antecedent risk factors hypertension, dyslipidemia, and cigarette smoking; 4) dose prescribed to the region of vascular disease; and 5) the use of chemotherapy (Table 1). Further information regarding the vascular disease obtained included the sites of vascular disease, presenting signs and symptoms, treatment, and final outcome.
Twenty-one patients (15 Hodgkin disease, 4 non-Hodgkin lymphoma, 2 seminoma) were identified who developed stenosis in a total of 41 noncoronary arterial sites after moderate dose irradiation. The median prescribed dose delivered to the central axis was 44 Gy (range, 30–50). Before 1991 a technique described by Kaplan was used at Stanford to keep uniformity of the total dose to within 5% of the prescribed dose to the midplane of the irradiated region, including the neck and supraclavicular regions. Dose uniformity to within 5% of the prescribed dose throughout the field was achieved by using partial field blocking and adjusting the number of treatments slightly upward for thicker areas or downward for thinner areas. The neck point was defined as just below the thyroid cartilage and the supraclavicular point was defined as the midclavicular line just superior to the clavicle.14 All but one of the patients was treated with the partial field blocking technique. The other patient was treated with a tissue compensator, which also kept the uniformity of the dose to within 5% of the prescribed dose. The median age at diagnosis of vessel disease was 54 years and the median time from therapy to event was 15 years (range, 1.3–40.0 yrs). Antecedent risk factors for atherosclerosis were prevalent in this group and included hypertension (52.4%), dyslipidemia (42.9%), and cigarette smoking (28.6%). Only three patients did not have any of these three risk factors.
Among 12 patients with carotid arterial disease, seven presented with neurologic symptoms, including three fatal cerebrovascular accidents, two strokes that resulted in hemiparesis, and one patient with amaurosis fugax. The pertinent clinical data, including age at vascular disease, dose, interval time period, presentation, surgical intervention, and outcome from vascular disease are shown in Tables 2 and 3. Tables 2 and 3 are divided among patients presenting with vascular disease before the age of 60 years as opposed to those after 60, as the causal relationship of vascular disease to radiation is less certain with age because of the natural formation of atherosclerosis with age. None of the five patients who had carotid disease identified by auscultation of a bruit has developed adverse neurologic sequelae. One of these five patients was treated with endarterectomy on one carotid artery and a subclavian bypass on the contralateral carotid, another was treated with endarterectomy, and the remaining three are being treated nonsurgically with medical management.
Table 2. Clinical Data for Patients Diagnosed with Vascular Disease before Age 60
Postprandial abdominal pain diagnosed as intestinal angina proved to be another ominous presentation of vascular disease, with two of three patients ultimately dying from bowel infarction. One of the three had occlusion of the left renal artery, superior mesenteric artery, and celiac axis (Fig. 1) and presented with postprandial pain, malabsorption, and diarrhea and weight loss. The second patient who also presented with postprandial pain was found to have diffuse disease involving the abdominal aorta, left renal artery, right common iliac artery, inferior mesenteric artery, and superior mesenteric artery on magnetic resonance (MR) angiogram (Fig. 2).
A 74-year-old man who presented with new onset hypertension and leg claudication on workup was found to have bilateral renal and left common iliac arterial stenosis (Fig. 3). Three other patients had symptoms of claudication and all three are doing well with medical management or stent placement. One patient developed chest pain and on subsequent cardiac catheterization was found to have right coronary artery disease as well as left subclavian stenosis. Finally, two men found to have subclavian stenosis presented very differently with differential blood pressure in the upper extremities in one patient and axillary edema in the other (Fig. 4).
The 21 patients in the study had a total of 41 sites of disease. Ten patients had multiple sites of disease within the radiation field. The most common site of vascular disease was the carotid artery, accounting for 34% of all the sites, and the second most common site was the subclavian artery, comprising 15% of all sites.
There were a total of five deaths that could be attributed to the vascular disease: three as a result of stroke and the other two from bowel infarction. Again, all five of the patients who had disease identified by auscultation of a bruit did not undergo any adverse clinical event. Twelve of the patients had their vascular disease treated without any surgical intervention, whereas the remaining nine patients had surgical procedures in the form of endarterectomy, bypass graft, angioplasty, and/or stent placement. The most common surgical procedure was endarterectomy, accounting for 35% of all the procedures.
Injury to small blood vessels has been identified as a fairly common factor in the pathogenesis of late radiation injury.4 Although large arteries appear to be the least affected by ionizing radiation, they can still produce dramatic effects. The degree of radiation vasculopathy depends on both the interval from treatment and the dose delivered.4 Most cases of radiation-related noncoronary vascular disease reported in the literature arose in individuals exposed to higher doses of radiation than those used for lymphoma or seminoma.6–9 The patients we reviewed show that noncoronary vascular disease appears to be a sporadic development many years after irradiation even in the intermediate dose range (30–50 Gy). Five patients died as a result of the vascular disease, with three deaths caused by stroke and the other two by small bowel infarction. A retrospective study of 910 patients showed a trend toward an increased incidence of stroke after neck irradiation in patients with Hodgkin disease, non-Hodgkin lymphoma, or head and neck carcinoma.13
The three general categories of radiation injury to large vessels are myointimal proliferation causing narrowing of the lumen, thrombosis, and rupture.2 The radiation injury to the patients studied here appears to result from myointimal proliferation. This consists of deposition of myofibroblasts, histiocytes, collagen, and fibrin resulting in fibrous intimal plaques, which lead to narrowing of the lumen.2
Similar to Hull et al.,11 there appears to be two subgroups of radiation survivors among our cohort of patients who developed noncoronary atherosclerotic vascular disease. The first group includes those patients with a diagnosis of vascular stenosis at an atypically young age of younger than 60 years old. There were 12 patients in this group, with a median age at diagnosis of vascular disease of 50 years. This group also had a longer latency period to event (median, 16.9 yrs) and featured unusual clinical presentations such as postprandial pain, axillary edema, and differential blood pressure in the upper extremities. This first group also appeared to have more adverse outcomes, with 50% of the patients having either death or stroke as a consequence of their vascular disease as opposed to 22% for the latter group of patients developing atherosclerotic vessel disease at a more typical age of older than 60 years. The second group with a median age of 70 years comprises the other nine patients who developed atherosclerotic vessel disease at a more typical age of older than 60 years. This second group had a shorter interval to development of vascular disease (median, 9.1 yrs) and more common presentations of vascular disease such as claudication. In fact, eight of the nine patients in the group diagnosed over the age of 60 presented with the more typical presentations of stroke, bruit, or claudication.
Disease occurring at an advanced age or a short interval between treatment and diagnosis of vascular disease suggests some preexisting atherosclerosis, and therefore make the impact of irradiation uncertain. Distinguishing radiation-induced vasculopathy from spontaneous atherosclerosis can be difficult. The cases more likely to be radiation-induced include those where vascular disease is limited to the radiation field, occurs at a younger age than expected, has a long interval from treatment to vasculopathy, and have lesions found in atypical locations. One patient was treated at the age of 5 and vascular hypoplasia may have been a contributing cause in this case.
The vast majority of our patients (86%) had at least one predisposing risk factor for development of atherosclerosis. These antecedent risk factors, along with the normal aging process, make less clear the amount of impact that radiation can have on developing atherosclerosis. Although it is clear that radiation contributes to the development of atherosclerosis, it is unclear whether it directly causes atherosclerosis or rather accelerates the process. An animal study looking at aortic arteriosclerosis in dogs after localized radiation suggested that the lesions at irradiated sites were usually more severe than nonirradiated sites, but otherwise very similar.5 It has been shown that a high cholesterol diet in rabbits acts synergistically with irradiation on the development of coronary artery disease.15 Radiation may just be an additional risk factor among patients who already have a significant risk of atherosclerosis.
Therefore, because radiation may act synergistically with other atherogenic factors, physicians should recognize the necessity to encourage modification of behavior and diet for patients treated with radiation. Furthermore, while following these patients physicians should be aware of the increased long-term risk for large vessel disease postirradiation and be cognizant of the varied ways in which large vessel stenosis may present.
These patients were not systematically evaluated for renal artery stenosis. Given the high prevalence of hypertension, perhaps this should be considered when patients develop hypertension after paraaortic or abdominal irradiation. Incorporating careful auscultation of the neck, upper chest, and abdomen for bruits in follow-up evaluation of irradiated patients may identify individuals who are at risk for adverse vascular events. The potential benefit from using auscultation for a bruit is suggested by the finding that all five patients in our study who presented with a bruit did not encounter any adverse neurologic sequelae with treatment. Also, physicians should consider vascular sources for odd gastrointestinal symptoms in individuals who had prior abdominal irradiation. There has been no prospective evaluation of carotid Doppler screening for people 5 years out from neck irradiation. Whether or not statins affect risk for postirradiation stenosis is unknown. Other strategies to decrease the risk of late effects may be to decrease the dose and/or volume of radiation with the use of chemotherapy. This has been the predominant approach at Stanford. Early stage Hodgkin disease is treated with risk-adapted Stanford V chemotherapy and radiation confined to a dose of 20 Gy.
A limitation of this study was the varied follow-up in the three different diagnostic groups, as the follow-up was more extensive for the Hodgkin disease group. Many of the Hodgkin disease patients were on protocols resulting in more extensive follow-up. In addition, the cases were identified by computer search of reported complication or cause of death and were not felt to be adequate to estimate the incidence or prevalence of this problem.
Although the techniques used at the time of treatment of most of the patients in this study were different from those currently used, these findings are still relevant today because use of a partial field blocking technique before 1991 kept the uniformity of the dose to within 5% of the prescribed dose in the midplane throughout the irradiated region.14 This is particularly important for the patients described here, because the actual dose to a region may be higher or lower depending on the thickness of the region. For example, the actual dose to the carotid and to a lesser degree the subclavian artery can be higher than the prescribed dose. The partial field blocking technique, originally described by Kaplan,14 was used for all but one patient in this study. Since 1991, tissue compensators have been used to achieve dose homogeneity to within 5% of the prescribed dose. In addition, a gap was calculated for adjacent fields, and a midline posterior spinal cord block was inserted from the superior border of inverted-Y or spade fields down to the lower border of the first lumbar vertebra after 20 Gy to prevent overdosing the field junction.14 Therefore, the reported prescribed doses are accurate to within 5% throughout the fields to which the patients in this study were treated.
Major arterial stenosis may occur as a long-term consequence of even modest dose irradiation. We hope that increased awareness among physicians of this potentially serious complication and its potential for unusual presentations, such as postprandial abdominal pain and axillary edema, may lead to more preventative measures and earlier diagnosis of radiation-induced large vessel disease.