Im ag i n g o f C an c er Th er ap y– Induced Central Nervous System Toxi ci ty Jörg Dietrich,
a, MD, PhD *,
Joshua P. Klein,
MD, PhD
b,
*
KEYWORDS Neurotoxicity Cancer Chemotherapy Radiation therapy Central nervous system Imaging KEY POINTS Cancer treatment–related neurotoxicity can occur in patients with both central nervous system and non-central nervous system cancers, and clinical symptoms are nonspecific. On structural imaging techniques, such as computed tomography and magnetic resonance imaging, distinguishing volume gain versus volume loss can be helpful in differentiating edema, inflammation, and tumor growth from gliosis, necrosis, and atrophy. Comparison of a current imaging study with recent and more remote prior imaging is crucial for recognizing subtle changes that occur over time. Radiation necrosis, leukoencephalopathy, hydrocephalus, and ischemic and hemorrhagic vascular events can occur with highly variable delay following treatment. Perfusion imaging and positron emission tomography can potentially help differentiate tumor progression versus necrosis and leukoencephalopathy, and functional magnetic resonance imaging can be helpful in surgical planning. Recent advanced imaging research studies in patients with cancer have provided novel insights into the cause of cognitive impairment and other neurotoxic syndromes in patients with cancer.
INTRODUCTION
Cancer therapies can cause a wide range of acute and delayed treatment complications involving the central nervous system (CNS),1,2 causing significant morbidity and mortality. Notably, these syndromes occur not only in patients with brain tumors but
a Division of Neuro-Oncology, Department of Neurology, Massachusetts General Hospital Cancer Center, Center for Regenerative Medicine, Harvard Medical School, 55 Fruit Street, Yawkey 9E, Boston, MA 02114, USA; b Department of Neurology, Brigham and Women’s Hospital, Harvard Medical School, Room AB-124, 75 Francis Street, Boston, MA 02115, USA * Corresponding author. E-mail addresses: [email protected]; [email protected]
Neurol Clin 32 (2014) 147–157 http://dx.doi.org/10.1016/j.ncl.2013.07.004 neurologic.theclinics.com 0733-8619/14/$ – see front matter Ó 2014 Elsevier Inc. All rights reserved.
148
Dietrich & Klein
also in patients treated for cancer outside the CNS, such as systemic lymphoma, breast, and lung cancer.3 With improved survival rates in patients with cancer and more aggressive and combined treatment modalities, neurologic treatment complications have been observed with increasing frequency. The clinical presentation of patients experiencing neurotoxicity is commonly nonspecific. Therefore, the diagnosis of neurotoxic syndromes often poses a major challenge to the treating physician. However, recognition of treatment-related neurologic complications is critically important to avoid unnecessary procedures, such as brain biopsy and lumbar puncture, and because symptoms may be confused with metastatic disease, tumor progression, paraneoplastic disorders, and infections of the CNS. Neuroimaging, including computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET), is an important diagnostic tool in patient evaluation and guidance of patient management. Moreover, advanced neuroimaging techniques with resting state and diffusion tensor MRI and functional MRI (fMRI) have most recently provided compelling evidence that both structural and functional brain changes occur in a substantial number of patients with cancer treated with chemotherapy and radiation. In this review various imaging techniques and modalities that can be used in patients with cancer with suspected neurotoxic syndromes are highlighted and advantages and limitations of each imaging modality in the context of classical neurotoxic syndromes encountered in patients with cancer are discussed. CT IN THE ASSESSMENT OF CANCER THERAPY–ASSOCIATED NEUROTOXIC SYNDROMES
CT is a technique that measures the attenuation of X-rays by different substances. Higher density materials absorb more x-rays than lower density materials and thus appear brighter (whiter). Two-dimensional images are reconstructed to display cross-sectional anatomy, and these individual images can be stacked so that one can view sequential images and appreciate the 3-dimensional contours of normal and abnormal structures. CT images can be acquired in any plane; for the brain, images are typically acquired in horizontal slices from the skull base to vertex. Slice thickness is variable, but most often approximately 5 mm per slice. Reformatted images in orthogonal planes (ie, coronal and sagittal) can be very useful in characterizing lesions. CT contrast is iodine-based; enhancement following contrast administration is seen in vascular structures and at sites of disruption of the blood-brain barrier. The appearance of tumor on CT, as well as the effects of its treatment, depends on many variables. The contents of the tumor and the presence of surrounding edema, hemorrhage, necrosis, and postsurgical changes attenuate radiographs to a different extent. The degree of attenuation is measured on a scale of Hounsfield units (HU), which ranges from 1000 (air) to 0 (water) to 11000 (dense bone or metal). Intracranially, cerebrospinal fluid (CSF) measures about 15 HU; white matter measures 20 to 30 HU, and gray matter measures 35 to 45 (HU). Fat, as in myelin, does not strongly attenuate radiographs and has an HU range of 30 to 70; thus, white matter appears darker than gray matter. Vasogenic edema in tissue appears dark on CT because water has a lower Hounsfield unit value than normal brain tissue. Vasogenic edema will most often be associated with local or regional mass effect. Subtle evidence of mass effect includes effacement of cortical sulci and compression of ventricles. With more severe mass effect, loss of gray-white matter differentiation and herniation of brain tissue is seen.
Neurotoxicity of Cancer Therapy
Gliotic and necrotic tissue also appears hypodense compared with normal brain. Differentiation between vasogenic edema and gliosis or necrosis relies on assessment of whether there is associated volume gain (ie, mass effect from edema) versus volume loss (ie, tissue destruction). These 2 opposing effects are often seen on the same image and cannot easily be distinguished (Fig. 1). Comparison to prior scans, if available, can help assess for interval volume gain or loss. To further complicate matters, CT cannot differentiate edema clearly from infiltrating tumors such as low-grade glioma or lymphoma. Hydrocephalus is an important imaging finding to recognize and occurs in 3 forms. With obstruction of ventricular outflow (noncommunicating hydrocephalus) or obstruction of CSF drainage into the venous circulation via arachnoid granulations (communicating hydrocephalus), there is elevated intracranial pressure. With cerebral atrophy and gliosis, ex vacuo ventricular dilatation occurs, and intracranial pressure may or may not be elevated. On CT, ventricular expansion with effacement of overlying cortical sulci is suggestive of noncommunicating hydrocephalus, and the most
Fig. 1. A 64-year-old patient with adenocystic carcinoma of the left mastoid underwent subtotal tumor resection and subsequent proton beam radiation. Several years later, axial brain MRI showed extensive abnormal T2 hyperintensity extending throughout the white matter of the left hemisphere (A) with irregular enhancement within the left temporal lobe (D). Several months later, the patient suffered a seizure and repeat imaging showed interval development of hydrocephalus, seen here as expansion of the right lateral ventricle (B). There was no change in the pattern or extent of enhancement at this time (E). Placement of a ventricular shunt relieved the hydrocephalus as seen on MRI several weeks later (C). Again, there was no change in the pattern or extent of enhancement (F). The imaging findings are consistent with radiation necrosis and leukoencephalopathy, rather than tumor progression.
149
150
Dietrich & Klein
common sites of obstruction are where the conduits of intraventricular CSF flow are narrowest (ie, the cerebral aqueduct). Ventriculomegaly in proportion to widening of the overlying cortical sulci is more suggestive of an ex vacuo effect. In patients who have undergone combinations of surgical resection, chemotherapy, and radiation, the combined and opposing elements of volume loss and edema can produce various types of hydrocephalus (see Fig. 1). As always, careful comparison with prior scans and correlation with clinical symptoms are essential. Despite these limitations, there are, in general, 3 settings in which imaging with CT may be a useful modality for patients with underlying cancer or known brain tumors: (1) with an abrupt change in neurologic examination, a CT can quickly and accurately assess for acute hemorrhage, herniation, or hydrocephalus; (2) if there is tumor extension into bone or if a tumor lies adjacent to bone, CT can be helpful to precisely delineate the margins of involvement and to assess for bony destruction (Fig. 2); (3) if a patient cannot undergo MRI because of metallic devices or some other reason, CT
Fig. 2. A 27-year-old patient with a history of childhood acute lymphoblastic leukemia treated with methotrexate-based chemotherapy and later with prophylactic cranial radiation presented with a painless bump on his right parietal scalp more than 15 years after initial treatment. CT revealed a focus of hyperostotic irregular bone in the right parietal skull. (A) Axial T1 MRI before (B) and after (C) gadolinium contrast administration revealed a round and homogenously enhancing mass arising from the right parietal dura with a dural tail. Axial T2-FLAIR (fluid attenuation inversion recovery) MRI (D) showed displacement of underlying right parietal cortex, with abnormal T2 hyperintensity in the subcortical white matter. Hyperintensity on diffusion-weighted MRI (E) and corresponding hypointensity on apparent diffusion coefficient maps (F), consistent with reduced diffusivity, is a marker of the hypercellularity of the mass, which was found to be a World Health Organization II meningioma and was treated with resection and adjuvant radiation. Meningiomas can arise as a complication from remote radiation therapy.4
Neurotoxicity of Cancer Therapy
with contrast and supplementary techniques such as CT perfusion and PET can be used to characterize a tumor and its treatment response. CT is highly sensitive for detecting hemorrhage, and acute hemorrhage and thrombus will appear hyperdense, typically at 60 to 80 HU. Blood becomes less hyperdense as it ages. Acute ischemia and infarction produce loss of gray-to-white matter differentiation and edema of involved tissue. Embolic or thrombotic occlusions of arteries and arterioles will most often produce wedge-shaped regions of hypodensity in corresponding arterial vascular distributions. Microvascular infarcts as seen with advanced atherosclerosis and lipohyalinosis following cranial radiation can occur in brain areas supplied by these tiny arteries, most often the basal ganglia and pons. Ischemic lesions disrupt the blood-brain barrier and will often enhance with contrast. Likewise, radiation necrosis is also commonly associated with abnormal contrast enhancement. This enhancement can be easily mistaken for tumor progression or recurrence and other imaging modalities can help differentiate these entities. Nonenhancing diffuse hypodensity within the cerebral white matter may indicate leukoencephalopathy and is better evaluated on MRI. Venous hypertension resulting from obstruction of a cortical vein or venous sinus can produce a venous infarct, which does not respect arterial vascular distributions and often has a disproportionate amount of edema and sometimes hemorrhage. MRI IN THE ASSESSMENT OF CANCER THERAPY–ASSOCIATED NEUROTOXIC SYNDROMES
MRI is obtained by aligning the natural atomic rotations of water molecules in the body within a strong magnetic field and superimposing brief radiofrequency pulses, to momentarily disalign their rotations. The manner in which molecules recover to their former alignment depends on the constituency of the tissue in which they are found and is what produces signal that can be reconstructed into images. For assessing the neurotoxic effects of cancer therapy, MRI is usually preferred over CT, because MRI is more sensitive to detecting changes both in the microenvironment surrounding a resection cavity and in brain structures that may be remote from a tumor but that have been exposed to radiation or chemotherapy. With MRI, the effects of cancer therapy can be assessed both structurally and functionally. Routine structural MRI essentially includes T1-weighted, T2-weighted, diffusion-weighted, and susceptibility-weighted sequences. With structural MRI, the principle of volume gain versus volume loss is the same as described for CT (Fig. 3). On T1-weighted sequences, white matter appears hyperintense (whiter) to gray matter because fat causes T1 shortening. Edema appears hypointense to normal white matter due to T1 prolongation of water within tissue. On T2-weighted sequences, white matter appears hypointense to gray matter because fat causes faster T2 signal decay. Edema appears hyperintense to normal white matter due to T2 prolongation of water within tissue (Fig. 4). Abnormal T2 prolongation is also seen in radiation-induced leukoencephalopathy and gliosis (see Figs. 1 and 3), which are commonly seen as delayed complication of CNS radiation therapy months to years after treatment.5–7 Diffusion-weighted imaging measures the ability of water molecules to diffuse freely within tissue. Restricted diffusion is seen in several scenarios including acute cytotoxic edema (ie, ischemic stroke), hypercellular tumors, and abscesses. Vasogenic edema, leukoencephalopathy,2 and gliosis are more often associated with elevated diffusivity, the “T2 shine-through” phenomenon. The appearance of hemorrhage on MRI depends on the state of decomposition of hemoglobin8 and can be
151
152
Dietrich & Klein
Fig. 3. A 66-year-old patient with a history of carcinoid lung cancer treated with lobectomy and adjuvant radiation and cisplatin-based chemotherapy, as well as prophylactic cranial radiation, presented several years later with progressive gait difficulties and cognitive decline. Coronal T2/FLAIR MRI at the time of cranial radiation (A) and at the time of re-presentation (B) show interval development of extensive abnormal T2 hyperintensity throughout the bi-hemispheric white matter, with loss of white matter volume and ex vacuo ventricular dilatation. There was no abnormal enhancement. The imaging findings are consistent with delayed leukoencephalopathy.
seen with tumor necrosis, radiation necrosis, primary hemorrhage, and hemorrhagic infarction (arterial and venous). Specialized sequences such as gradient echo and susceptibility-weighted imaging can detect microhemorrhages that are too small to be seen on T1-weighted and T2-weighted sequences. The added sensitivity of the gradient echo and susceptibility-weighted imaging sequences is due to “blooming artifact,” which causes microhemorrhages to appear slightly larger than their true size.
Fig. 4. A 58-year-old patient with multiple myeloma was treated with cyclophosphamide and dexamethasone, and after recurrence, underwent additional chemotherapy followed by autologous stem cell transplantation. The patient continued to have disease progression and was started on a proteasome inhibitor. Two days after infusion, the patient developed a headache, then blurred vision, and then had a generalized seizure. Axial T2-FLAIR MRI revealed multiple foci of abnormal T2 hyperintensity within the bilateral frontoparietal (A, B) and occipital (C, D) subcortical white matter and overlying gray matter, with cortical sulcal effacement. There was no abnormal enhancement or hemorrhage. These findings are consistent with posterior reversible leukoencephalopathy syndrome.27,28 Of note, the patient was never found to be hypertensive and had no abrupt changes in blood pressure. The proteasome inhibitor was discontinued and symptoms improved over the next few weeks. Many cytotoxic immunosuppressant medications have been associated with posterior reversible leukoencephalopathy syndrome, including cyclosporin, tacrolimus, sirolimus, cisplatin, and bevacizumab.3
Neurotoxicity of Cancer Therapy
Like CT, contrast enhancement on MRI is seen in vascular structures and at sites of blood-brain barrier disruption. Increase or decrease in enhancement following specific therapies does not necessarily reflect tumor progression, which may complicate the interpretation of imaging findings in patients with cancer; instead, they may simply reflect changes in blood-brain barrier permeability, which is affected by both tumor and treatment. The phenomena of “pseudoprogression” and “pseudoresponse” have been described to bring attention to this potentially confounding aspect of image interpretation, which has been studied most extensively in glioblastoma.9,10 The principles of recognizing hydrocephalus are the same on MRI as on CT (see Figs. 1 and 3). With noncommunicating hydrocephalus, the additional finding of transependymal flow of CSF appears as smooth “caps” of uniform and confluent abnormal T2 hyperintensity within the periventricular white matter. Also, like CT, radiation necrosis can induce variable edema as well as abnormal enhancement (Figs. 5 and 6). FUNCTIONAL IMAGING IN THE ASSESSMENT OF CANCER THERAPY–ASSOCIATED NEUROTOXIC SYNDROMES
Advanced imaging modalities, such as perfusion imaging, PET, and fMRI, are being used with increasing frequency to assess tumors and their response to treatment. Perfusion imaging can be performed with both CT and MRI. The technique usually requires an infusion of contrast and measurement of changes in enhancement of tissue as contrast transits through it. A time-intensity curve is created, from which the interdependent measurements of blood volume, blood flow, and transit time can be calculated. Highly vascular tumors will demonstrate elevated blood volume, whereas necrotic or ischemic tissue will demonstrate reduced blood flow (ischemic) versus reduced blood volume (infarcted or necrotic).11 This technique is particularly useful for assessing abnormal posttreatment enhancement that can occur as a result of both tumor recurrence or radiation necrosis. The PET technique can be used to study cellular metabolism and pharmacology by generating a map of positron-emitting-radioisotope-labeled biomolecules in living tissues. In oncology, the radioisotope [18F]fluoro-deoxyglucose (FDG) is commonly used to measure glucose uptake as a surrogate marker of cellular metabolism. Highly metabolically active tissues such as the cerebral cortex have intrinsically high glucose
Fig. 5. A 65-year-old patient with melanoma metastatic to the left frontal lobe underwent resection and proton beam stereotactic radiosurgery and brachytherapy. The patient developed progressive loss of vision in the left and right eye several years after radiation therapy. Coronal T2 MRI revealed abnormal hyperintensity in the left greater than right optic chiasm as well as the pituitary infundibulum and inferior basal ganglia on the left (A). Coronal (B) and axial (C) T1 MRI following gadolinium contrast administration showed abnormal enhancement of the left optic chiasm and nerve. These findings are consistent with radiation-induced optic neuropathy.29 The patient was initially treated with dexamethasone and later bevacizumab, with clinical and radiographic stabilization.
153
154
Dietrich & Klein
Fig. 6. A 23-year-old patient underwent resection of a right frontal lobe glioblastoma with adjuvant standard radiation and temozolomide. Subsequent brain imaging 3 to 4 years later revealed a large right frontal lobe resection cavity with interval development of multifocal areas of abnormal T2 hyperintensity (A) in the left frontal lobe with irregular enhancement on T1 postgadolinium MRI (B). There was no evidence of abnormal reduced diffusivity (hypointensity) on apparent diffusion coefficient maps (C), and no abnormal elevated blood volume (hyperintensity) on MR perfusion (D). These lesions have remained largely stable on subsequent imaging over 2 to 3 years, consistent with delayed radiation necrosis, rather than tumor recurrence.
utilization. Rapidly growing tumors can demonstrate elevated glucose utilization as well. PET can be coregistered with structural images acquired from CT and MRI, and in this setting, the metabolic state of a mass lesion can be assessed. Hypermetabolism suggests malignancy, whereas hypometabolism may represent necrosis or nonneoplastic inflammation.12 fMRI is used in surgical planning for tumors in or adjacent to eloquent areas of brain; fMRI signal generation is based on the concept of neurovascular coupling, whereby elevated local cerebral blood flow occurs in response to elevated activity of a population of neurons. Blood-oxygen–dependent changes are detected by comparing blood flow before and during a specific task. Motor, sensory, visual, and language areas of the brain can often be precisely localized, and the proximity of these areas to a tumor can be helpful in assessing the risk of tumor resection. EMERGING IMAGING TECHNIQUES AND FUTURE DIRECTIONS IN CANCER THERAPY– ASSOCIATED NEUROTOXICITY
Several recent research studies using MRI, fMRI, and PET in the assessment of cancer therapy–associated neurotoxicity have provided evidence that structural and functional CNS changes occur in a significant number of patients with cancer treated with systemic chemotherapy.13–16 For instance, structural MRI studies in patients with breast cancer treated with chemotherapy have revealed decreased regional volumes of gray and white matter, including the prefrontal and parahippocampal areas.17 Consistent with these findings, recent MRI studies in a similar patient population identified reduction in overall brain volume,18 and specifically in frontal, temporal, and cerebellar cortex in patients examined longitudinally before and after chemotherapy.19 Using diffusion tensor imaging, an MRI-based technique that makes vector-based, or directional, measurements of reduced diffusivity to reconstruct maps of white matter tracts, white matter damage related to systemic chemotherapy has been demonstrated.13,20–22 In addition, a small number of recent fMRI studies supports the notion that systemic chemotherapy-induced cognitive impairment and decreased executive function correlates with regionally altered brain function.23–25 A functional imaging study using [15O] water and [18F]FDG-PET in a cohort of breast cancer survivors treated with tamoxifen-based chemotherapy 5 to 10 years before
Neurotoxicity of Cancer Therapy
identified alterations in frontocortical, cerebellar, and basal ganglia metabolism.26 Reduced resting glucose metabolism in inferior-frontal brain regions correlated with impaired short-term memory function. However, the lack of imaging data before chemotherapy was one of the major limitations of this study. As this field of investigation is rapidly expanding, ongoing and future imaging research studies are likely to reveal novel biomarkers of neurotoxicity that will allow identification of patients with highest risk to develop neurotoxic syndromes as the consequence of their cancer treatment. Collectively, advanced imaging studies have identified both structural and functional brain changes as a consequence of cancer therapy, therefore challenging the previous dogma that the adult CNS is largely resistant to the toxic effects of systemic chemotherapy. With a growing understanding of the effects of chemotherapy and radiation in the brain, future clinical trials in patients with cancer are therefore expected to increasingly incorporate advanced imaging modalities along with other biomarker studies (eg, genetic, epigenetic, and metabolic) to assess and predict treatment-related neurotoxicity. In addition, steadily expanding knowledge about imaging biomarkers of neurotoxicity will facilitate identification and validation of neuroprotective strategies with the overall goal of improved management and quality of life of patients with cancer. ACKNOWLEDGMENTS
The authors gratefully acknowledge the support of the American Academy of Neurology Foundation (J. Dietrich), the American Cancer Society (J. Dietrich), and the Stephen E. and Catherine Pappas Center Research Foundation (J. Dietrich). J. Dietrich is a Fellow of the Clinical Investigator Training Program (CITP) at Harvard Medical School. REFERENCES
1. Dietrich J, Monje M, Wefel J, et al. Clinical patterns and biological correlates of cognitive dysfunction associated with cancer therapy. Oncologist 2008;13: 1285–95. 2. Arrillaga-Romany I, Dietrich J. Imaging findings in cancer therapy-associated neurotoxicity. Semin Neurol 2012;32:476–86. 3. Dietrich J, Wen P. Neurologic complications of chemotherapy. In: Schiff D, Kesari S, Wen P, editors. Cancer neurology in clinical practice. Totowa (NJ): Humana Press Inc; 2008. p. 287–326. 4. Galloway TJ, Indelicato DJ, Amdur RJ, et al. Second tumors in pediatric patients treated with radiotherapy to the central nervous system. Am J Clin Oncol 2012;35: 279–83. 5. Wang YX, King AD, Zhou H, et al. Evolution of radiation-induced brain injury: MR imaging-based study. Radiology 2010;254:210–8. 6. Rahmathulla G, Marko NF, Weil RJ. Cerebral radiation necrosis: a review of the pathobiology, diagnosis and management considerations. J Clin Neurosci 2013;20(4):485–502. 7. Yoshii Y. Pathological review of late cerebral radionecrosis. Brain Tumor Pathol 2008;25:51–8. 8. Bradley WG Jr. MR appearance of hemorrhage in the brain. Radiology 1993;189: 15–26. 9. Hygino da Cruz LC Jr, Rodriguez I, Domingues RC, et al. Pseudoprogression and pseudoresponse: imaging challenges in the assessment of posttreatment glioma. AJNR Am J Neuroradiol 2011;32:1978–85.
155
156
Dietrich & Klein
10. Tanaka S, Louis DN, Curry WT, et al. Diagnostic and therapeutic avenues for glioblastoma: no longer a dead end? Nat Rev Clin Oncol 2013;10:14–26. 11. Hu LS, Baxter LC, Smith KA, et al. Relative cerebral blood volume values to differentiate high-grade glioma recurrence from posttreatment radiation effect: direct correlation between image-guided tissue histopathology and localized dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging measurements. AJNR Am J Neuroradiol 2009;30:552–8. 12. Chao ST, Suh JH, Raja S, et al. The sensitivity and specificity of FDG PET in distinguishing recurrent brain tumor from radionecrosis in patients treated with stereotactic radiosurgery. Int J Cancer 2001;96:191–7. 13. Deprez S, Amant F, Smeets A, et al. Longitudinal assessment of chemotherapyinduced structural changes in cerebral white matter and its correlation with impaired cognitive functioning. J Clin Oncol 2012;30:274–81. 14. Koppelmans V, Breteler MM, Boogerd W, et al. Neuropsychological performance in survivors of breast cancer more than 20 years after adjuvant chemotherapy. J Clin Oncol 2012;30:1080–6. 15. McDonald BC, Conroy SK, Ahles TA, et al. Alterations in brain activation during working memory processing associated with breast cancer and treatment: a prospective functional magnetic resonance imaging study. J Clin Oncol 2012; 30:2500–8. 16. McDonald BC, Conroy SK, Smith DJ, et al. Frontal gray matter reduction after breast cancer chemotherapy and association with executive symptoms: a replication and extension study. Brain Behav Immun 2013;30(Suppl): S117–25. 17. Inagaki M, Yoshikawa E, Matsuoka Y, et al. Smaller regional volumes of brain gray and white matter demonstrated in breast cancer survivors exposed to adjuvant chemotherapy. Cancer 2007;109:146–56. 18. Koppelmans V, de Ruiter MB, van der Lijn F, et al. Global and focal brain volume in long-term breast cancer survivors exposed to adjuvant chemotherapy. Breast Cancer Res Treat 2012;132:1099–106. 19. McDonald BC, Conroy SK, Ahles TA, et al. Gray matter reduction associated with systemic chemotherapy for breast cancer: a prospective MRI study. Breast Cancer Res Treat 2010;123:819–28. 20. Abraham J, Haut MW, Moran MT, et al. Adjuvant chemotherapy for breast cancer: effects on cerebral white matter seen in diffusion tensor imaging. Clin Breast Cancer 2008;8:88–91. 21. de Ruiter MB, Reneman L, Boogerd W, et al. Cerebral hyporesponsiveness and cognitive impairment 10 years after chemotherapy for breast cancer. Hum Brain Mapp 2011;32:1206–19. 22. Deprez S, Amant F, Yigit R, et al. Chemotherapy-induced structural changes in cerebral white matter and its correlation with impaired cognitive functioning in breast cancer patients. Hum Brain Mapp 2011;32:480–93. 23. Ferguson RJ, McDonald BC, Saykin AJ, et al. Brain structure and function differences in monozygotic twins: possible effects of breast cancer chemotherapy. J Clin Oncol 2007;25:3866–70. 24. Kesler SR, Bennett FC, Mahaffey ML, et al. Regional brain activation during verbal declarative memory in metastatic breast cancer. Clin Cancer Res 2009; 15:6665–73. 25. Kesler SR, Kent JS, O’Hara R. Prefrontal cortex and executive function impairments in primary breast cancer. Arch Neurol 2011;68:1447–53.
Neurotoxicity of Cancer Therapy
26. Silverman DH, Dy CJ, Castellon SA, et al. Altered frontocortical, cerebellar, and basal ganglia activity in adjuvant-treated breast cancer survivors 5-10 years after chemotherapy. Breast Cancer Res Treat 2007;103:303–11. 27. Bartynski WS, Boardman JF. Distinct imaging patterns and lesion distribution in posterior reversible encephalopathy syndrome. AJNR Am J Neuroradiol 2007; 28:1320–7. 28. Marinella MA, Markert RJ. Reversible posterior leucoencephalopathy syndrome associated with anticancer drugs. Intern Med J 2009;39:826–34. 29. Zhao Z, Lan Y, Bai S, et al. Late-onset radiation-induced optic neuropathy after radiotherapy for nasopharyngeal carcinoma. J Clin Neurosci 2013;20(5):702–6.
157
a, MD, PhD *,
Joshua P. Klein,
MD, PhD
b,
*
KEYWORDS Neurotoxicity Cancer Chemotherapy Radiation therapy Central nervous system Imaging KEY POINTS Cancer treatment–related neurotoxicity can occur in patients with both central nervous system and non-central nervous system cancers, and clinical symptoms are nonspecific. On structural imaging techniques, such as computed tomography and magnetic resonance imaging, distinguishing volume gain versus volume loss can be helpful in differentiating edema, inflammation, and tumor growth from gliosis, necrosis, and atrophy. Comparison of a current imaging study with recent and more remote prior imaging is crucial for recognizing subtle changes that occur over time. Radiation necrosis, leukoencephalopathy, hydrocephalus, and ischemic and hemorrhagic vascular events can occur with highly variable delay following treatment. Perfusion imaging and positron emission tomography can potentially help differentiate tumor progression versus necrosis and leukoencephalopathy, and functional magnetic resonance imaging can be helpful in surgical planning. Recent advanced imaging research studies in patients with cancer have provided novel insights into the cause of cognitive impairment and other neurotoxic syndromes in patients with cancer.
INTRODUCTION
Cancer therapies can cause a wide range of acute and delayed treatment complications involving the central nervous system (CNS),1,2 causing significant morbidity and mortality. Notably, these syndromes occur not only in patients with brain tumors but
a Division of Neuro-Oncology, Department of Neurology, Massachusetts General Hospital Cancer Center, Center for Regenerative Medicine, Harvard Medical School, 55 Fruit Street, Yawkey 9E, Boston, MA 02114, USA; b Department of Neurology, Brigham and Women’s Hospital, Harvard Medical School, Room AB-124, 75 Francis Street, Boston, MA 02115, USA * Corresponding author. E-mail addresses: [email protected]; [email protected]
Neurol Clin 32 (2014) 147–157 http://dx.doi.org/10.1016/j.ncl.2013.07.004 neurologic.theclinics.com 0733-8619/14/$ – see front matter Ó 2014 Elsevier Inc. All rights reserved.
148
Dietrich & Klein
also in patients treated for cancer outside the CNS, such as systemic lymphoma, breast, and lung cancer.3 With improved survival rates in patients with cancer and more aggressive and combined treatment modalities, neurologic treatment complications have been observed with increasing frequency. The clinical presentation of patients experiencing neurotoxicity is commonly nonspecific. Therefore, the diagnosis of neurotoxic syndromes often poses a major challenge to the treating physician. However, recognition of treatment-related neurologic complications is critically important to avoid unnecessary procedures, such as brain biopsy and lumbar puncture, and because symptoms may be confused with metastatic disease, tumor progression, paraneoplastic disorders, and infections of the CNS. Neuroimaging, including computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET), is an important diagnostic tool in patient evaluation and guidance of patient management. Moreover, advanced neuroimaging techniques with resting state and diffusion tensor MRI and functional MRI (fMRI) have most recently provided compelling evidence that both structural and functional brain changes occur in a substantial number of patients with cancer treated with chemotherapy and radiation. In this review various imaging techniques and modalities that can be used in patients with cancer with suspected neurotoxic syndromes are highlighted and advantages and limitations of each imaging modality in the context of classical neurotoxic syndromes encountered in patients with cancer are discussed. CT IN THE ASSESSMENT OF CANCER THERAPY–ASSOCIATED NEUROTOXIC SYNDROMES
CT is a technique that measures the attenuation of X-rays by different substances. Higher density materials absorb more x-rays than lower density materials and thus appear brighter (whiter). Two-dimensional images are reconstructed to display cross-sectional anatomy, and these individual images can be stacked so that one can view sequential images and appreciate the 3-dimensional contours of normal and abnormal structures. CT images can be acquired in any plane; for the brain, images are typically acquired in horizontal slices from the skull base to vertex. Slice thickness is variable, but most often approximately 5 mm per slice. Reformatted images in orthogonal planes (ie, coronal and sagittal) can be very useful in characterizing lesions. CT contrast is iodine-based; enhancement following contrast administration is seen in vascular structures and at sites of disruption of the blood-brain barrier. The appearance of tumor on CT, as well as the effects of its treatment, depends on many variables. The contents of the tumor and the presence of surrounding edema, hemorrhage, necrosis, and postsurgical changes attenuate radiographs to a different extent. The degree of attenuation is measured on a scale of Hounsfield units (HU), which ranges from 1000 (air) to 0 (water) to 11000 (dense bone or metal). Intracranially, cerebrospinal fluid (CSF) measures about 15 HU; white matter measures 20 to 30 HU, and gray matter measures 35 to 45 (HU). Fat, as in myelin, does not strongly attenuate radiographs and has an HU range of 30 to 70; thus, white matter appears darker than gray matter. Vasogenic edema in tissue appears dark on CT because water has a lower Hounsfield unit value than normal brain tissue. Vasogenic edema will most often be associated with local or regional mass effect. Subtle evidence of mass effect includes effacement of cortical sulci and compression of ventricles. With more severe mass effect, loss of gray-white matter differentiation and herniation of brain tissue is seen.
Neurotoxicity of Cancer Therapy
Gliotic and necrotic tissue also appears hypodense compared with normal brain. Differentiation between vasogenic edema and gliosis or necrosis relies on assessment of whether there is associated volume gain (ie, mass effect from edema) versus volume loss (ie, tissue destruction). These 2 opposing effects are often seen on the same image and cannot easily be distinguished (Fig. 1). Comparison to prior scans, if available, can help assess for interval volume gain or loss. To further complicate matters, CT cannot differentiate edema clearly from infiltrating tumors such as low-grade glioma or lymphoma. Hydrocephalus is an important imaging finding to recognize and occurs in 3 forms. With obstruction of ventricular outflow (noncommunicating hydrocephalus) or obstruction of CSF drainage into the venous circulation via arachnoid granulations (communicating hydrocephalus), there is elevated intracranial pressure. With cerebral atrophy and gliosis, ex vacuo ventricular dilatation occurs, and intracranial pressure may or may not be elevated. On CT, ventricular expansion with effacement of overlying cortical sulci is suggestive of noncommunicating hydrocephalus, and the most
Fig. 1. A 64-year-old patient with adenocystic carcinoma of the left mastoid underwent subtotal tumor resection and subsequent proton beam radiation. Several years later, axial brain MRI showed extensive abnormal T2 hyperintensity extending throughout the white matter of the left hemisphere (A) with irregular enhancement within the left temporal lobe (D). Several months later, the patient suffered a seizure and repeat imaging showed interval development of hydrocephalus, seen here as expansion of the right lateral ventricle (B). There was no change in the pattern or extent of enhancement at this time (E). Placement of a ventricular shunt relieved the hydrocephalus as seen on MRI several weeks later (C). Again, there was no change in the pattern or extent of enhancement (F). The imaging findings are consistent with radiation necrosis and leukoencephalopathy, rather than tumor progression.
149
150
Dietrich & Klein
common sites of obstruction are where the conduits of intraventricular CSF flow are narrowest (ie, the cerebral aqueduct). Ventriculomegaly in proportion to widening of the overlying cortical sulci is more suggestive of an ex vacuo effect. In patients who have undergone combinations of surgical resection, chemotherapy, and radiation, the combined and opposing elements of volume loss and edema can produce various types of hydrocephalus (see Fig. 1). As always, careful comparison with prior scans and correlation with clinical symptoms are essential. Despite these limitations, there are, in general, 3 settings in which imaging with CT may be a useful modality for patients with underlying cancer or known brain tumors: (1) with an abrupt change in neurologic examination, a CT can quickly and accurately assess for acute hemorrhage, herniation, or hydrocephalus; (2) if there is tumor extension into bone or if a tumor lies adjacent to bone, CT can be helpful to precisely delineate the margins of involvement and to assess for bony destruction (Fig. 2); (3) if a patient cannot undergo MRI because of metallic devices or some other reason, CT
Fig. 2. A 27-year-old patient with a history of childhood acute lymphoblastic leukemia treated with methotrexate-based chemotherapy and later with prophylactic cranial radiation presented with a painless bump on his right parietal scalp more than 15 years after initial treatment. CT revealed a focus of hyperostotic irregular bone in the right parietal skull. (A) Axial T1 MRI before (B) and after (C) gadolinium contrast administration revealed a round and homogenously enhancing mass arising from the right parietal dura with a dural tail. Axial T2-FLAIR (fluid attenuation inversion recovery) MRI (D) showed displacement of underlying right parietal cortex, with abnormal T2 hyperintensity in the subcortical white matter. Hyperintensity on diffusion-weighted MRI (E) and corresponding hypointensity on apparent diffusion coefficient maps (F), consistent with reduced diffusivity, is a marker of the hypercellularity of the mass, which was found to be a World Health Organization II meningioma and was treated with resection and adjuvant radiation. Meningiomas can arise as a complication from remote radiation therapy.4
Neurotoxicity of Cancer Therapy
with contrast and supplementary techniques such as CT perfusion and PET can be used to characterize a tumor and its treatment response. CT is highly sensitive for detecting hemorrhage, and acute hemorrhage and thrombus will appear hyperdense, typically at 60 to 80 HU. Blood becomes less hyperdense as it ages. Acute ischemia and infarction produce loss of gray-to-white matter differentiation and edema of involved tissue. Embolic or thrombotic occlusions of arteries and arterioles will most often produce wedge-shaped regions of hypodensity in corresponding arterial vascular distributions. Microvascular infarcts as seen with advanced atherosclerosis and lipohyalinosis following cranial radiation can occur in brain areas supplied by these tiny arteries, most often the basal ganglia and pons. Ischemic lesions disrupt the blood-brain barrier and will often enhance with contrast. Likewise, radiation necrosis is also commonly associated with abnormal contrast enhancement. This enhancement can be easily mistaken for tumor progression or recurrence and other imaging modalities can help differentiate these entities. Nonenhancing diffuse hypodensity within the cerebral white matter may indicate leukoencephalopathy and is better evaluated on MRI. Venous hypertension resulting from obstruction of a cortical vein or venous sinus can produce a venous infarct, which does not respect arterial vascular distributions and often has a disproportionate amount of edema and sometimes hemorrhage. MRI IN THE ASSESSMENT OF CANCER THERAPY–ASSOCIATED NEUROTOXIC SYNDROMES
MRI is obtained by aligning the natural atomic rotations of water molecules in the body within a strong magnetic field and superimposing brief radiofrequency pulses, to momentarily disalign their rotations. The manner in which molecules recover to their former alignment depends on the constituency of the tissue in which they are found and is what produces signal that can be reconstructed into images. For assessing the neurotoxic effects of cancer therapy, MRI is usually preferred over CT, because MRI is more sensitive to detecting changes both in the microenvironment surrounding a resection cavity and in brain structures that may be remote from a tumor but that have been exposed to radiation or chemotherapy. With MRI, the effects of cancer therapy can be assessed both structurally and functionally. Routine structural MRI essentially includes T1-weighted, T2-weighted, diffusion-weighted, and susceptibility-weighted sequences. With structural MRI, the principle of volume gain versus volume loss is the same as described for CT (Fig. 3). On T1-weighted sequences, white matter appears hyperintense (whiter) to gray matter because fat causes T1 shortening. Edema appears hypointense to normal white matter due to T1 prolongation of water within tissue. On T2-weighted sequences, white matter appears hypointense to gray matter because fat causes faster T2 signal decay. Edema appears hyperintense to normal white matter due to T2 prolongation of water within tissue (Fig. 4). Abnormal T2 prolongation is also seen in radiation-induced leukoencephalopathy and gliosis (see Figs. 1 and 3), which are commonly seen as delayed complication of CNS radiation therapy months to years after treatment.5–7 Diffusion-weighted imaging measures the ability of water molecules to diffuse freely within tissue. Restricted diffusion is seen in several scenarios including acute cytotoxic edema (ie, ischemic stroke), hypercellular tumors, and abscesses. Vasogenic edema, leukoencephalopathy,2 and gliosis are more often associated with elevated diffusivity, the “T2 shine-through” phenomenon. The appearance of hemorrhage on MRI depends on the state of decomposition of hemoglobin8 and can be
151
152
Dietrich & Klein
Fig. 3. A 66-year-old patient with a history of carcinoid lung cancer treated with lobectomy and adjuvant radiation and cisplatin-based chemotherapy, as well as prophylactic cranial radiation, presented several years later with progressive gait difficulties and cognitive decline. Coronal T2/FLAIR MRI at the time of cranial radiation (A) and at the time of re-presentation (B) show interval development of extensive abnormal T2 hyperintensity throughout the bi-hemispheric white matter, with loss of white matter volume and ex vacuo ventricular dilatation. There was no abnormal enhancement. The imaging findings are consistent with delayed leukoencephalopathy.
seen with tumor necrosis, radiation necrosis, primary hemorrhage, and hemorrhagic infarction (arterial and venous). Specialized sequences such as gradient echo and susceptibility-weighted imaging can detect microhemorrhages that are too small to be seen on T1-weighted and T2-weighted sequences. The added sensitivity of the gradient echo and susceptibility-weighted imaging sequences is due to “blooming artifact,” which causes microhemorrhages to appear slightly larger than their true size.
Fig. 4. A 58-year-old patient with multiple myeloma was treated with cyclophosphamide and dexamethasone, and after recurrence, underwent additional chemotherapy followed by autologous stem cell transplantation. The patient continued to have disease progression and was started on a proteasome inhibitor. Two days after infusion, the patient developed a headache, then blurred vision, and then had a generalized seizure. Axial T2-FLAIR MRI revealed multiple foci of abnormal T2 hyperintensity within the bilateral frontoparietal (A, B) and occipital (C, D) subcortical white matter and overlying gray matter, with cortical sulcal effacement. There was no abnormal enhancement or hemorrhage. These findings are consistent with posterior reversible leukoencephalopathy syndrome.27,28 Of note, the patient was never found to be hypertensive and had no abrupt changes in blood pressure. The proteasome inhibitor was discontinued and symptoms improved over the next few weeks. Many cytotoxic immunosuppressant medications have been associated with posterior reversible leukoencephalopathy syndrome, including cyclosporin, tacrolimus, sirolimus, cisplatin, and bevacizumab.3
Neurotoxicity of Cancer Therapy
Like CT, contrast enhancement on MRI is seen in vascular structures and at sites of blood-brain barrier disruption. Increase or decrease in enhancement following specific therapies does not necessarily reflect tumor progression, which may complicate the interpretation of imaging findings in patients with cancer; instead, they may simply reflect changes in blood-brain barrier permeability, which is affected by both tumor and treatment. The phenomena of “pseudoprogression” and “pseudoresponse” have been described to bring attention to this potentially confounding aspect of image interpretation, which has been studied most extensively in glioblastoma.9,10 The principles of recognizing hydrocephalus are the same on MRI as on CT (see Figs. 1 and 3). With noncommunicating hydrocephalus, the additional finding of transependymal flow of CSF appears as smooth “caps” of uniform and confluent abnormal T2 hyperintensity within the periventricular white matter. Also, like CT, radiation necrosis can induce variable edema as well as abnormal enhancement (Figs. 5 and 6). FUNCTIONAL IMAGING IN THE ASSESSMENT OF CANCER THERAPY–ASSOCIATED NEUROTOXIC SYNDROMES
Advanced imaging modalities, such as perfusion imaging, PET, and fMRI, are being used with increasing frequency to assess tumors and their response to treatment. Perfusion imaging can be performed with both CT and MRI. The technique usually requires an infusion of contrast and measurement of changes in enhancement of tissue as contrast transits through it. A time-intensity curve is created, from which the interdependent measurements of blood volume, blood flow, and transit time can be calculated. Highly vascular tumors will demonstrate elevated blood volume, whereas necrotic or ischemic tissue will demonstrate reduced blood flow (ischemic) versus reduced blood volume (infarcted or necrotic).11 This technique is particularly useful for assessing abnormal posttreatment enhancement that can occur as a result of both tumor recurrence or radiation necrosis. The PET technique can be used to study cellular metabolism and pharmacology by generating a map of positron-emitting-radioisotope-labeled biomolecules in living tissues. In oncology, the radioisotope [18F]fluoro-deoxyglucose (FDG) is commonly used to measure glucose uptake as a surrogate marker of cellular metabolism. Highly metabolically active tissues such as the cerebral cortex have intrinsically high glucose
Fig. 5. A 65-year-old patient with melanoma metastatic to the left frontal lobe underwent resection and proton beam stereotactic radiosurgery and brachytherapy. The patient developed progressive loss of vision in the left and right eye several years after radiation therapy. Coronal T2 MRI revealed abnormal hyperintensity in the left greater than right optic chiasm as well as the pituitary infundibulum and inferior basal ganglia on the left (A). Coronal (B) and axial (C) T1 MRI following gadolinium contrast administration showed abnormal enhancement of the left optic chiasm and nerve. These findings are consistent with radiation-induced optic neuropathy.29 The patient was initially treated with dexamethasone and later bevacizumab, with clinical and radiographic stabilization.
153
154
Dietrich & Klein
Fig. 6. A 23-year-old patient underwent resection of a right frontal lobe glioblastoma with adjuvant standard radiation and temozolomide. Subsequent brain imaging 3 to 4 years later revealed a large right frontal lobe resection cavity with interval development of multifocal areas of abnormal T2 hyperintensity (A) in the left frontal lobe with irregular enhancement on T1 postgadolinium MRI (B). There was no evidence of abnormal reduced diffusivity (hypointensity) on apparent diffusion coefficient maps (C), and no abnormal elevated blood volume (hyperintensity) on MR perfusion (D). These lesions have remained largely stable on subsequent imaging over 2 to 3 years, consistent with delayed radiation necrosis, rather than tumor recurrence.
utilization. Rapidly growing tumors can demonstrate elevated glucose utilization as well. PET can be coregistered with structural images acquired from CT and MRI, and in this setting, the metabolic state of a mass lesion can be assessed. Hypermetabolism suggests malignancy, whereas hypometabolism may represent necrosis or nonneoplastic inflammation.12 fMRI is used in surgical planning for tumors in or adjacent to eloquent areas of brain; fMRI signal generation is based on the concept of neurovascular coupling, whereby elevated local cerebral blood flow occurs in response to elevated activity of a population of neurons. Blood-oxygen–dependent changes are detected by comparing blood flow before and during a specific task. Motor, sensory, visual, and language areas of the brain can often be precisely localized, and the proximity of these areas to a tumor can be helpful in assessing the risk of tumor resection. EMERGING IMAGING TECHNIQUES AND FUTURE DIRECTIONS IN CANCER THERAPY– ASSOCIATED NEUROTOXICITY
Several recent research studies using MRI, fMRI, and PET in the assessment of cancer therapy–associated neurotoxicity have provided evidence that structural and functional CNS changes occur in a significant number of patients with cancer treated with systemic chemotherapy.13–16 For instance, structural MRI studies in patients with breast cancer treated with chemotherapy have revealed decreased regional volumes of gray and white matter, including the prefrontal and parahippocampal areas.17 Consistent with these findings, recent MRI studies in a similar patient population identified reduction in overall brain volume,18 and specifically in frontal, temporal, and cerebellar cortex in patients examined longitudinally before and after chemotherapy.19 Using diffusion tensor imaging, an MRI-based technique that makes vector-based, or directional, measurements of reduced diffusivity to reconstruct maps of white matter tracts, white matter damage related to systemic chemotherapy has been demonstrated.13,20–22 In addition, a small number of recent fMRI studies supports the notion that systemic chemotherapy-induced cognitive impairment and decreased executive function correlates with regionally altered brain function.23–25 A functional imaging study using [15O] water and [18F]FDG-PET in a cohort of breast cancer survivors treated with tamoxifen-based chemotherapy 5 to 10 years before
Neurotoxicity of Cancer Therapy
identified alterations in frontocortical, cerebellar, and basal ganglia metabolism.26 Reduced resting glucose metabolism in inferior-frontal brain regions correlated with impaired short-term memory function. However, the lack of imaging data before chemotherapy was one of the major limitations of this study. As this field of investigation is rapidly expanding, ongoing and future imaging research studies are likely to reveal novel biomarkers of neurotoxicity that will allow identification of patients with highest risk to develop neurotoxic syndromes as the consequence of their cancer treatment. Collectively, advanced imaging studies have identified both structural and functional brain changes as a consequence of cancer therapy, therefore challenging the previous dogma that the adult CNS is largely resistant to the toxic effects of systemic chemotherapy. With a growing understanding of the effects of chemotherapy and radiation in the brain, future clinical trials in patients with cancer are therefore expected to increasingly incorporate advanced imaging modalities along with other biomarker studies (eg, genetic, epigenetic, and metabolic) to assess and predict treatment-related neurotoxicity. In addition, steadily expanding knowledge about imaging biomarkers of neurotoxicity will facilitate identification and validation of neuroprotective strategies with the overall goal of improved management and quality of life of patients with cancer. ACKNOWLEDGMENTS
The authors gratefully acknowledge the support of the American Academy of Neurology Foundation (J. Dietrich), the American Cancer Society (J. Dietrich), and the Stephen E. and Catherine Pappas Center Research Foundation (J. Dietrich). J. Dietrich is a Fellow of the Clinical Investigator Training Program (CITP) at Harvard Medical School. REFERENCES
1. Dietrich J, Monje M, Wefel J, et al. Clinical patterns and biological correlates of cognitive dysfunction associated with cancer therapy. Oncologist 2008;13: 1285–95. 2. Arrillaga-Romany I, Dietrich J. Imaging findings in cancer therapy-associated neurotoxicity. Semin Neurol 2012;32:476–86. 3. Dietrich J, Wen P. Neurologic complications of chemotherapy. In: Schiff D, Kesari S, Wen P, editors. Cancer neurology in clinical practice. Totowa (NJ): Humana Press Inc; 2008. p. 287–326. 4. Galloway TJ, Indelicato DJ, Amdur RJ, et al. Second tumors in pediatric patients treated with radiotherapy to the central nervous system. Am J Clin Oncol 2012;35: 279–83. 5. Wang YX, King AD, Zhou H, et al. Evolution of radiation-induced brain injury: MR imaging-based study. Radiology 2010;254:210–8. 6. Rahmathulla G, Marko NF, Weil RJ. Cerebral radiation necrosis: a review of the pathobiology, diagnosis and management considerations. J Clin Neurosci 2013;20(4):485–502. 7. Yoshii Y. Pathological review of late cerebral radionecrosis. Brain Tumor Pathol 2008;25:51–8. 8. Bradley WG Jr. MR appearance of hemorrhage in the brain. Radiology 1993;189: 15–26. 9. Hygino da Cruz LC Jr, Rodriguez I, Domingues RC, et al. Pseudoprogression and pseudoresponse: imaging challenges in the assessment of posttreatment glioma. AJNR Am J Neuroradiol 2011;32:1978–85.
155
156
Dietrich & Klein
10. Tanaka S, Louis DN, Curry WT, et al. Diagnostic and therapeutic avenues for glioblastoma: no longer a dead end? Nat Rev Clin Oncol 2013;10:14–26. 11. Hu LS, Baxter LC, Smith KA, et al. Relative cerebral blood volume values to differentiate high-grade glioma recurrence from posttreatment radiation effect: direct correlation between image-guided tissue histopathology and localized dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging measurements. AJNR Am J Neuroradiol 2009;30:552–8. 12. Chao ST, Suh JH, Raja S, et al. The sensitivity and specificity of FDG PET in distinguishing recurrent brain tumor from radionecrosis in patients treated with stereotactic radiosurgery. Int J Cancer 2001;96:191–7. 13. Deprez S, Amant F, Smeets A, et al. Longitudinal assessment of chemotherapyinduced structural changes in cerebral white matter and its correlation with impaired cognitive functioning. J Clin Oncol 2012;30:274–81. 14. Koppelmans V, Breteler MM, Boogerd W, et al. Neuropsychological performance in survivors of breast cancer more than 20 years after adjuvant chemotherapy. J Clin Oncol 2012;30:1080–6. 15. McDonald BC, Conroy SK, Ahles TA, et al. Alterations in brain activation during working memory processing associated with breast cancer and treatment: a prospective functional magnetic resonance imaging study. J Clin Oncol 2012; 30:2500–8. 16. McDonald BC, Conroy SK, Smith DJ, et al. Frontal gray matter reduction after breast cancer chemotherapy and association with executive symptoms: a replication and extension study. Brain Behav Immun 2013;30(Suppl): S117–25. 17. Inagaki M, Yoshikawa E, Matsuoka Y, et al. Smaller regional volumes of brain gray and white matter demonstrated in breast cancer survivors exposed to adjuvant chemotherapy. Cancer 2007;109:146–56. 18. Koppelmans V, de Ruiter MB, van der Lijn F, et al. Global and focal brain volume in long-term breast cancer survivors exposed to adjuvant chemotherapy. Breast Cancer Res Treat 2012;132:1099–106. 19. McDonald BC, Conroy SK, Ahles TA, et al. Gray matter reduction associated with systemic chemotherapy for breast cancer: a prospective MRI study. Breast Cancer Res Treat 2010;123:819–28. 20. Abraham J, Haut MW, Moran MT, et al. Adjuvant chemotherapy for breast cancer: effects on cerebral white matter seen in diffusion tensor imaging. Clin Breast Cancer 2008;8:88–91. 21. de Ruiter MB, Reneman L, Boogerd W, et al. Cerebral hyporesponsiveness and cognitive impairment 10 years after chemotherapy for breast cancer. Hum Brain Mapp 2011;32:1206–19. 22. Deprez S, Amant F, Yigit R, et al. Chemotherapy-induced structural changes in cerebral white matter and its correlation with impaired cognitive functioning in breast cancer patients. Hum Brain Mapp 2011;32:480–93. 23. Ferguson RJ, McDonald BC, Saykin AJ, et al. Brain structure and function differences in monozygotic twins: possible effects of breast cancer chemotherapy. J Clin Oncol 2007;25:3866–70. 24. Kesler SR, Bennett FC, Mahaffey ML, et al. Regional brain activation during verbal declarative memory in metastatic breast cancer. Clin Cancer Res 2009; 15:6665–73. 25. Kesler SR, Kent JS, O’Hara R. Prefrontal cortex and executive function impairments in primary breast cancer. Arch Neurol 2011;68:1447–53.
Neurotoxicity of Cancer Therapy
26. Silverman DH, Dy CJ, Castellon SA, et al. Altered frontocortical, cerebellar, and basal ganglia activity in adjuvant-treated breast cancer survivors 5-10 years after chemotherapy. Breast Cancer Res Treat 2007;103:303–11. 27. Bartynski WS, Boardman JF. Distinct imaging patterns and lesion distribution in posterior reversible encephalopathy syndrome. AJNR Am J Neuroradiol 2007; 28:1320–7. 28. Marinella MA, Markert RJ. Reversible posterior leucoencephalopathy syndrome associated with anticancer drugs. Intern Med J 2009;39:826–34. 29. Zhao Z, Lan Y, Bai S, et al. Late-onset radiation-induced optic neuropathy after radiotherapy for nasopharyngeal carcinoma. J Clin Neurosci 2013;20(5):702–6.
157
