Nuclear accidents and acts of terrorism have the potential to expose thousands of people to high-dose total-body iradiation (TBI). Those who survive the acute radiation syndrome are at risk of developing chronic, degenerative radiation-induced injuries [delayed effects of acute radiation (DEARE)] that may negatively affect quality of life. A growing body of literature suggests that the brain may be vulnerable to radiation injury at survivable doses, yet the long-term consequences of high-dose TBI on the adult brain are unclear. Herein we report the occurrence of lesions consistent with cerebrovascular injury, detected by susceptibility-weighted magnetic resonance imaging (MRI), in a cohort of non-human primate [(NHP); rhesus macaque, Macaca mulatta] long-term survivors of high-dose TBI (1.1–8.5 Gy). Animals were monitored longitudinally with brain MRI (approximately once every three years). Susceptibility-weighted images (SWI) were reviewed for hypointensities (cerebral microbleeds and/or focal necrosis). SWI hypointensities were noted in 13% of irradiated NHP; lesions were not observed in control animals. A prior history of exposure was correlated with an increased risk of developing a lesion detectable by MRI (P = 0.003). Twelve of 16 animals had at least one brain lesion present at the time of the first MRI evaluation; a subset of animals (n = 7) developed new lesions during the surveillance period (3.7–11.3 years postirradiation). Lesions occurred with a predilection for white matter and the gray–white matter junction. The majority of animals with lesions had one to three SWI hypointensities, but some animals had multifocal disease (n = 2). Histopathologic evaluation of deceased animals within the cohort (n = 3) revealed malformation of the cerebral vasculature and remodeling of the blood vessel walls. There was no association between comorbid diabetes mellitus or hypertension with SWI lesion status. These data suggest that long-term TBI survivors may be at risk of developing cerebrovascular injury years after irradiation.
Acts of nuclear terrorism and incidents involving nuclear power reactors have the capacity to expose hundreds of thousands of people to potentially harmful doses of radiation (1). Radiation exposure in these scenarios is presumed unavoidable. Patients who survive the acute radiation syndrome (ARS) and those receiving lower doses of radiation may be at risk of developing delayed effects of acute radiation exposure (DEARE), comprised of a myriad of degenerative and inflammatory conditions affecting multiple organs, and with negative effect on quality of life (2). Thus there is a profound need to understand the long-term health effects of such radiation exposures to enable adequate risk assessment and direct patient follow-up and management practices.
The brain is considered radioresistant, with single dose total-body irradiation (TBI) of 20–50 Gy required to induce acute radiation syndrome (ARS) of the central nervous system (3). However, a growing body of literature suggests that the brain may be vulnerable to injury at lower doses (4–18). Retrospective analyses of atomic bomb survivors indicate that low-dose in utero radiation exposure (≥0.06 Gy at 8–15 weeks' gestation and ≥0.28 Gy 16–25 weeks' gestation) may result in dose-related reduction in cognitive function, microcephaly and seizures (19–24). Neurocognitive impairment may also occur in children receiving myeloablative doses of TBI for hematopoietic cell transplantation (25–32). Studies evaluating the long-term ramifications of single-dose TBI on the adult brain have been limited; however, retrospective analyses of long-term radiation survivor cohorts indicate survivors may have increased risk of stroke and vascular encephalopathy (13–15).
Studies evaluating the long-term effects of radiation in humans are confounded by factors that include socioeconomic status, genetic variation, postirradiation survival interval and comorbidities such as cancer, hypertension and diabetes mellitus (33–35). These factors can be controlled for in rodent studies, and radiation-induced changes in the brain have been extensively characterized in rodents, including impairments in neurogenesis (36–38), changes in dendritic spine morphology and density (39–41), changes in neuroinflammation and ATP production (18), impaired glutamatergic neurotransmission (42–46), and increased astrocytic and microglial inflammation (47–55). However, rodents do not develop white matter necrosis after fractionated whole-brain irradiation (fWBI) (44, 56–58), and differences in neuroanatomy [(e.g., greater hippocampal relative brain volume, larger gray:white matter ratio) (59–61)] and limited inter-individual genetic variability (62) may limit the translatability of findings. In contrast to rodents, non-human primates have greater neuroanatomic and genomic homology to humans and more fully recapitulate the histopathologic features of late-delayed radiation-induced brain injury (RIBI) after fWBI (e.g., cerebrovascular and white matter injury) (63–68).
The Radiation Countermeasures Center of Research Excellence (RadCCORE) non-human primate Radiation Survivor Cohort (RSC) is a unique resource in which rhesus macaques that have survived high-dose TBI (1.14–8.5 Gy) are adopted and monitored longitudinally for the development of late-delayed sequelae of radiation exposure. In this cohort, differences in diet and environment are controlled for, and health status extensively characterized (e.g., cardiac function, metabolic status, immune function, cancer surveillance), which makes this cohort ideal for the study of late effects of TBI and interactions between comorbid radiation-induced conditions.
Although the long-term consequences of single-high-dose TBI on the adult brain are unclear, fWBI for the treatment of brain cancer may result in neuroinflammation, white matter necrosis, vascular injury and cognitive impairment (69–72). In affected patients, scattered foci of hemorrhage and necrosis consistent with small blood vessel injury are detectable on susceptibility-weighted (SWI) magnetic resonance images (MRI) (73–78). In the current study, we hypothesized that non-human primate long-term survivors of high-dose TBI would develop lesions on imaging that were similar to those observed after fWBI, although less severe and with a longer latency period.
MATERIALS AND METHODS
Magnetic resonance imaging (MRI) scans acquired between 2011 and 2018 were evaluated from 173 rhesus macaques, aged 2.8–18.3 years (median age: 6.5 years). Of these animals, 141 were males and 32 were females. In addition, 53 animals (50 males, 3 females; median age: 6.2 years, range: 3.5–18.3) had no prior history of irradiation and were either part of unrelated studies or adopted as control comparators and enrolled within the non-human primate Radiation Survivor Cohort (RSC). A total of 120 animals (91 males, 29 females; median age: 6.5 years, range: 2.8–16.5 years) were administered single-dose TBI (median dose: 6.6 Gy, range: 1.14–8.5 Gy) as part of separate studies at other institutions [Wake Forest University (Winston-Salem, NC); University of Maryland (College Park, MD), University of Illinois (Urbana-Champaign, IL); Armed Forces Radiobiology Research Institute (Bethesda, MD); Lovelace Respiratory Research Institute (Albuquerque, NM); and Citox Labs (Stillwell, KS)] and then adopted into the RSC for long-term surveillance for the sequelae of radiation exposure. Additional detail on irradiation methods, supportive care strategies, and acute effects for many animals donated to this cohort have been recently reported elsewhere (79–81), and are summarized in Supplementary Table S1 (277_rare-194-03-06_s01.pdf) ( https://doi.org/10.1667/RADE-20-00051.1.S1). Tabulated demographic data include age, radiation dose, dose rate and age at irradiation (Table 1).
Animals were housed in either stainless-steel cages in temperature- and humidity-controlled rooms with a 12:12 h light-dark schedule, or indoor-outdoor pens; animals were socially housed whenever possible. Diet was contingent upon the experimental study in which the animal was enrolled; animals were either fed commercially available monkey chow (Purina®, St. Louis, MO) or Typical American Primate diet (LabDiet 5L0P; Land O'Lakes Inc., St. Louis, MO] designed to approximate the macronutrients of a Western dietary profile. All animals received daily clinical assessment by trained veterinary personnel, including cage-side neurologic evaluation when indicated; all animals were neurologically normal. All procedures were approved by the Wake Forest University School of Medicine (WFUSM) Institutional Animal Care and Use Committee (IACUC) and performed in accordance with all state and federal animal welfare laws. WFUSM is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC).
Magnetic Resonance Imaging
Animals were sedated with ketamine HCl (15 mg/kg body weight, IM) and maintained on inhaled isoflurane (3% induction, 1.5% maintenance) in 100% oxygen anesthesia for the duration of the MRI procedure. Images were acquired on a 3.0-Tesla Siemens Skyra clinical magnetic resonance scanner (Malvern, PA), operating on a D13 platform with a maximum gradient field strength = 45 mT/m. Sequences were optimized for non-human primates and acquired using a dedicated, 8 channel, non-human primate radiofrequency coil (Rapid MR International, Columbus, OH) tuned to 127.7 MHz.
T1-weighted anatomic images were obtained using a 3D volumetric magnetization-prepared rapid acquisition with gradient echo (MPRAGE) sequence (TR = 2,700 ms; TE = 3.39 ms; TI = 880 ms; FA = 8 degrees; 160 slices, voxel dimension = 0.5 × 0.5 × 0.5 mm3). Brain lesions consistent with hemorrhage and necrosis were identified by susceptibility-weighted image (SWI) sequence (TR = 28 ms, TE = 20 ms, flip angle = 15 degrees, NEX = 2, with voxel size = 0.5 × 0.5 × 0.5 mm3). SWI and T1 sequences were co-registered, and the anatomic location of SWI-hypointense foci was determined from the corresponding T1 structural image. All lesion dimensions were calculated from T1 structural images. Lesion diameters were measured in three perpendicularly-oriented planes; volume was estimated assuming ellipsoid shape ( abc), where a, b and c represent the axial, coronal and sagittal diameters.
Archived data regarding comorbid conditions were retrieved from an internal animal medical records system and assessed for evidence of type II diabetes mellitus (T2DM) and hypertension. A diagnosis of T2DM was assigned by laboratory clinical veterinary personnel and was determined by clinical presentation including hyperglycemia, glucosuria or ketonuria, dyslipidemia, elevated hemoglobin A1C and/ or glucose tolerance test results. Hypertension was defined as systolic blood pressure >130 mmHg and/or diastolic blood pressure >80 mmHg; the average of the three most recent measurements of systolic and diastolic blood pressure for each animal were used.
All statistical analyses were completed using GraphPad Prism version 7.0 (La Jolla, CA). Animals were stratified by irradiation and SWI lesion status. Normality was assessed using Shapiro-Wilk test. Non-parametric data (age at irradiation, age at lesion diagnosis, dose, dose rate, and irradiation-to-diagnosis interval) were compared by Mann Whitney U. The age at time of assessment between nonirradiated, irradiated NSL (no significant lesions) and irradiated SWI groups was compared using Kruskal-Wallis one-way analysis of variance (ANOVA). Outliers for lesion number were determined by ROUT method (Q = 1%; GraphPad Prism). The correlation between radiation dose (Gy) and lesion volume was assessed using two-tailed Spearman correlation.
The associations between prior radiation exposure and brain lesions on MRI, and between comorbid health conditions (T2DM and hypertension) and brain lesions on MRI were assessed using Fisher's exact test and the odds ratio calculated by Woolf logit method. Significance for all analyses was set to P ≤ 0.05.
Susceptibility-Weighted Imaging (SWI) Hypointense Brain Lesions on MRI
Focal and multifocal T1-hypointense, SWI-hypointense brain lesions consistent with necrosis and/or cerebrovascular injury were noted in 13% of irradiated NHP and 0% of nonirradiated comparators (Fig. 1, Table 2). Prior TBI was associated with an increased risk of developing a lesion detectable by MRI (Fisher's exact, P = 0.003). No significant proprioceptive, locomotor or cranial nerve deficits were noted on cage-side neurologic evaluation.
Animals Receiving TBI are More Likely to Develop SWI Hypointense Lesions Detectable on MRI
Animals that developed brain lesions received higher doses of TBI (Fig. 2, Table 1) than animals that did not develop SWI-hypointense lesions; there was no statistically significant difference in dose rate. Twelve out of 16 (12/16) animals had at least one brain lesion present at the time of the first MRI evaluation; 7/16 animals developed nascent lesions during the study period (Fig. 3), of which 4/7 had no antecedent lesions. Nascent lesions (n = 11) developed 3.7–11.3 years postirradiation (median: 7.3 years) at 9.9–15.6 years of age (median: 11.3 years). There were no statistically significant differences in radiation dose or age at irradiation between animals that developed nascent lesions and those with antecedent lesions only.
There were 55 SWI-hypointense brain lesions across 16 animals; 14 out of 16 animals had 1–3 lesions and the remaining animals had 16 lesions each (n = 2) (Fig. 4A). No statistically significant relationship was found between radiation dose and SWI lesion number (Fig. 4B). Lesions occurred most frequently within the cerebrum (89%), with predilection for the occipital lobe (45%) (Table 3). Seventeen lesions were below the resolution capability of the T1-weighted sequence, precluding further neuroanatomic evaluation. Of the remaining lesions (38/55), the majority occurred within white matter or at the gray-white junction (50% and 29%, respectively); 21% occurred within gray matter. Lesions were equally distributed between hemispheres (right: n = 27; left: n = 28).
Neuroanatomic Distribution of SWI Hypointense Foci
SWI and T1-weighted images were aligned, and lesion volumes calculated from the corresponding T1-weighted structural image to circumvent overestimation of lesion size due to the blooming artifact inherent in SWI (82). The majority of lesions (47/55, 85%) were ≤1.0 × 1.0 × 1.0 mm3 (0.5 mm3), 17 of which were below the resolution capacity of the T1-weighted sequence and recorded with a lesion volume of 0 mm3. The median cumulative lesion volume per animal was 1.3 mm3 (range: 0–14.9 mm3). Longitudinal volumetric data were available for 26 lesions (n = 11 subjects); once present, the majority of lesions did not change in size over time (62%, 16/26 lesions), while 9 lesions increased in volume (n = 5 subjects). One lesion was no longer present on re-evaluation one year later. No statistically significant relationship was found between median or cumulative lesion volume and radiation dose (Gy) (P > 0.05).
The associations between comorbid health conditions (T2DM and hypertension) and brain lesion presence were individually assessed. No statistically significant association between T2DM or hypertension and brain lesion status was detected.
Correlative Gross and Histopathology
Of the 16 animals diagnosed with SWI-hypointense brain lesions, two animals (Table 4) were humanely euthanized due to illnesses unassociated with neurologic dysfunction. NHP1 was euthanized due to the presence of multiple neoplasms (chondrolipoma, renal carcinoma, and poorly-differentiated neuroendocrine neoplasia) and poor prognosis. There was no gross or histologic evidence of metastatic disease. NHP2 was euthanized due to persistent bacterial pneumonia despite antibiotic treatment and supportive care, concurrent heart murmur, and T2DM.
Demographics of Deceased Animals with SWI Hypointense Brain Lesions
In NHP1, a focal SWI hypointensity with corresponding 2.2 × 1.7 × 1.7 mm3 (3.3 mm3) T1 parenchymal defect was observed within the deep white matter of the caudal right occipital lobe at the time of first MRI assessment 5.8 years postirradiation (Fig. 5A and B). At reassessment 7.3 years postirradiation, the lesion had increased to 2.3 × 2.5 × 2.4 mm3 (7.2 mm3). At this time, a second SWI-hypointense focus with corresponding 5.9 × 1.2 × 0.9 mm3 (3.3 mm3) T1 parenchymal defect was found within the area between the right optic tract and globus pallidus. Gross evaluation of the occipital lobe revealed an irregular area of hemorrhage (Fig. 5C); histopathology revealed a focal vascular malformation resembling a cavernous hemangioma (Fig. 6A). There was scant hemorrhage in the ventral leptomeninges adjacent to the optic chiasm. An artery adjacent to the optic tract was segmentally, transmurally disrupted and replaced by aggregates of eosinophilic extracellular matrix and hemosiderin-laden macrophages (Fig. 6B). Distal segments were tortuous and redundant, with partial-to-complete occlusion of the vascular lumina by eosinophilic extracellular matrix and smooth muscle hyperplasia (Fig. 6C). White matter tracts adjacent to both vascular morphologic abnormalities were vacuolated, and there was axonal degeneration (swollen axons, dilated myelin sheathes and digestion chambers) within the cerebral peduncle.
In NHP2, no significant lesions were noted at the time of first MRI assessment 5.9 years postirradiation. At 6.8 years postirradiation, there was a focal SWI hypointensity with corresponding 0.4 × 0.3 × 0.3 mm3 (0.02 mm3) T1 parenchymal defect within the deep white matter of the left ventromedial occipital lobe. This lesion was present at reassessment 8.4 and 9.4 years postirradiation and did not change in size over time. The lesion could not be found grossly or histopathologically. A second SWI hypointensity with corresponding 0.4 × 0.4 × 0.4 mm3 (0.03 mm3) T1 parenchymal defect was noted within the white matter of the left prefrontal lobe at 8.4 and 9.4 years postirradiation. The lesion did not change in size over time. Grossly, there was focal hemorrhage within the prefrontal white matter. The corresponding histopathology evaluation revealed dilated tortuous vasculature with extensive perivascular extracellular matrix and chronic hemorrhage, morphologically consistent with radiation-induced cerebrovascular injury.
Cerebrovascular injury occurring without a premonitory SWI lesion was noted in one animal at the time of necropsy, after humane euthanasia necessitated by the occurrence of hepatocellular carcinoma (10.2 years of age, 5.4 years after 7.2 Gy TBI). On sectioned surface, there was regionally extensive congestion within the deep white matter of the right occipital lobe. Histopathology revealed a 4 × 5 mm2 focus in which the deep white matter was infiltrated by irregularly dilated small- to medium-caliber blood vessels (telangiectasia, Fig. 6D). There were no SWI-hypointense lesions within the occipital lobe on a routine surveillance MRI conducted six months prior to necropsy.
Herein we have documented the prevalence of brain lesions consistent with focal or multifocal vascular injury in a cohort of rhesus macaque long-term survivors of single, high-dose TBI. We also provide two case reports detailing the corresponding histopathologic findings. These data add to a growing body of literature which suggests that the brain is vulnerable to radiation injury at doses lower than the CNS-ARS threshold.
Magnetic resonance imaging is a non-invasive imaging technique used to evaluate the response of hydrogen atoms to magnetic fields and radiofrequency pulses to reconstruct structural and functional tissue parameters as visual data. Modulation of the applied radiofrequency pulses in specific “sequences” is used to assess different biological tissue characteristics. In the current study, T1-weighted MPRAGE sequences were used to acquire anatomic brain images and provide contrast between gray and white matter. SWI is extremely sensitive to the presence of diamagnetic (calcium) and paramagnetic (iron, hemosiderin) substances within the body (82, 83), which appear as foci of “blooming” susceptibility/hypointense artifact. Thus these sequences can be used to identify hemorrhage and necrosis with mineralization in affected subjects. The SWI sequence is routinely used in clinical practice to identify blood products and small-volume microhemorrhages in human patients, and SWI hypointensities have been reported in cancer patients receiving brain radiotherapy (73, 74, 77, 78, 84–88). Use of these non-invasive, in vivo imaging protocols allows for monitoring of the progression of brain pathology longitudinally, often years prior to death, and offers the potential to map MRI abnormalities to neurologic deficits.
As animals in this study are currently enrolled in a cohort undergoing long-term surveillance for health effects of prior radiation exposure, animals are not euthanized until declining health status necessitates humane euthanasia. Therefore, our analyses were primarily limited to the evaluation of findings on magnetic resonance images acquired in vivo. Histopathologic evaluation was limited to the two deceased animals noted to have SWI-hypointense lesions on MRI (14/16 are living). While caution should be used in extrapolating these results to the general population, the histopathologic appearance agrees with the imaging characteristics (73–75, 89, 90). Vascular malformations (cavernous hemangiomas) occur as a sequela of cerebral irradiation in humans (89, 91, 92). Despite previous categorization as “hemangiomas,” these focal structural abnormalities are not presumed due to proliferation of endothelium, and in keeping with the International Society for the Study of Vascular Anomalies (ISSVA) classification system, are more correctly classified as “vascular malformations” (93). The remaining cerebrovascular lesions observed in this study (perivascular extracellular matrix with acute and chronic hemorrhage) are similar to lesions observed in late-delayed RIBI after fWBI in non-human primates and humans (64, 67, 68, 70, 90, 94, 95). These cases suggest the etiopathogenesis of late-delayed RIBI involves vascular injury characterized by inappropriate remodeling and resultant structural abnormalities. Identification of cerebrovascular injury at necropsy in the absence of a corresponding lesion on brain imaging, as had occurred in one animal, suggests that some of these lesions may develop rapidly. Additional studies will follow, including comprehensive histopathologic evaluation of the remaining 14/16 animals.
In one case, a lesion noted on MRI could not be located on gross or histopathologic evaluation, despite use of the MRI data during the trimming process and stereological sectioning. The identification of brain lesions at the time of prosection is limited by small lesion size, the relatively large volume of the non-human primate brain, and low optical contrast between foci of necrosis and normal tissue.
Our study was limited to the evaluation of postirradiation MRI, as animals within the cohort are adopted from separate studies in which acquisition of preirradiation brain MRI was not a part of the experimental design. Hemorrhages develop in the brain parenchyma and subarachnoid spaces in mice within 30 days after LD50/30 TBI (9.5 Gy 60Co), and occur in association with thrombocytopenia (17, 18, 96). Concurrent neuroinflammation, increased cytokine production and deficits in ATP biogenesis (18) may also predispose the cerebral vasculature to increased permeability or vascular necrosis, leading to parenchymal hemorrhage during the acute phases of radiation injury. Consequently, congenital lesions and residual iron resultant from thrombocytopenia or endothelial injury during ARS cannot be excluded as the cause of the lesions present at the time of first MRI assessment; however, the development of additional lesions during the study period in seven animals (3.7–11.3 years postirradiation) suggests a component of ongoing (chronic) cerebrovascular injury. The affected animals in the current study were not thrombocytopenic at the time of MRI. The absence of similar foci on MRI in nonirradiated controls supports the hypothesis that these lesions are radiation-induced. While cognitive assessment within this cohort has been limited due to subject unwillingness to engage with the task, preliminary analyses suggest that some animals may have reduced cognitive flexibility (97). Radiation-induced temporal microbleeds in cancer patients are independently associated with likelihood of cognitive dysfunction (78). SWI lesions in this cohort occur in the absence of neurological signs detectable by cranial nerve and proprioception testing. Further research evaluating the physiological and cognitive ramifications, and progression of these lesions, is needed and will be the focus of additional studies.
T2DM and cardiovascular disease are late-delayed effects of TBI and have been observed in this cohort of macaques (98, 99). As these conditions also have effects on the cerebral vasculature in the absence of radiation exposure, we evaluated the association between comorbid T2DM and hypertension on SWI lesion occurrence. No statistically significant association between concurrent T2DM or hypertension and SWI-hypointense brain lesions was found.
In summary, we have characterized the imaging prevalence of brain lesions compatible with cerebrovascular injury in a cohort of rhesus macaque long-term survivors of high dose TBI. We have determined that prior radiation exposure is associated with the development of SWI-hypointense brain lesions and provide evidence that the brain may be susceptible to injury at lower doses of radiation than those required to induce the ARS of the central nervous system (20–50 Gy) (3). These data suggest that human populations exposed to high-dose TBI may benefit from long-term follow-up including MRI to assess cerebrovascular injury.
Table S1 (277_rare-194-03-06_s01.pdf). Irradiation procedures.
We acknowledge the efforts of the Cline Laboratory technical staff for their assistance in acquiring MRI images and providing care to animals within the RadCCORE RSC. This work was supported by NIH grant no. U19-AI67798 awarded to Nelson Chao, Duke University (JMC: primate core leader, Wake Forest School of Medicine). EGB was supported by NIH grant no. T35 OD 10946.