Chemotherapy: neurologic complications
May. 10, 2022
At vero eos et accusamus et iusto odio dignissimos ducimus qui blanditiis praesentium voluptatum deleniti atque corrupti quos dolores et quas.
Low-grade gliomas consist of WHO grade II primary CNS tumors with a glial lineage. These tumors, often described as “diffuse” and/or “infiltrating” typically present with seizures in otherwise healthy young adults. Low-grade gliomas have a better prognosis than anaplastic gliomas and advances in understanding of tumor genetics have reconfigured classification, improved prognostication, and helped identify patients who may respond more effectively to therapies. The authors summarize the basic biology, clinical features, and evolving treatment options for patients with low-grade gliomas.
• Low-grade gliomas account for approximately 15% of glial tumors in adults.
• Most patients present with seizures, often with a history of focal seizures persisting for weeks or months.
• Tumor genetics and molecular pathology are now a pillar of the most recent WHO classification.
• Common practice is maximal safe resection. Other treatment options include radiation therapy, chemotherapy, or deferred intervention.
This article will focus on WHO grade II CNS tumors of primary glial origin, specifically infiltrating gliomas. Under the current WHO classification scheme, this includes diffuse astrocytoma (IDH-mutant, IDH-wild type, and NOS), oligodendroglioma (IDH mutant and 1p/19q co-deleted, and NOS), and oligoastrocytoma NOS (23). Other “low-grade” tumors with glial elements, such as pilocytic astrocytoma (WHO grade I), as well as tumors with mixed glial/neuronal elements, including ganglioglioma, dysembryoblastic neuroepithelial tumor, and central neurocytoma, are separate entities and are not discussed here.
The current WHO 2016 classification requires demonstration of IDH1 or IDH2 mutation status for low-grade gliomas of all histologic subtypes as well as demonstration of 1p/19q codeletion status for oligodendroglial histologic subtypes. Notably, the WHO classification discourages the diagnosis of tumors as oligoastrocytoma or “mixed glioma”, although the entity of oligoastrocytoma NOS is still described in detail. These declarations are new from the 2007 WHO classification, and represent a nosological “reshuffling” of certain glioma patients, both low- and high-grade. In clinical practice, this has translated into a relative blurring of previously hardened delineations between oligodendroglial, astrocytic, and mixed histologic subtypes.
The typical low-grade glioma patient presents between their early twenties and mid-forties (33). The majority of patients present with seizures, which can manifest as focal or generalized (38). Compared to higher-grade glial tumors, it is less common for low-grade gliomas to present with focal neurologic deficits or signs of increased intracranial pressure such as positional headache or blurred vision (12). Lesions deeper in the brain, further from cortical structures, are more likely to present focally (09). In the present era of CT and MR imaging, more often tumors are discovered incidentally on studies performed for trauma, migraine, dizziness, etc.
The course of disease varies widely based on tumor location and volume and other prognostic features such as age, surgical quality, and molecular subgroup, as discussed below.
A neurooncologist will combine a clinical evaluation with imaging, histopathology, and tumor molecular analysis to prognosticate an individual low-grade glioma patient. Historical datasets, trial data, and large analyses have helped identify well-known prognostic markers, which can now be combined with key tumor biomarkers to glean more precise estimates of outcomes. This is particularly important for low-grade gliomas because survival can vary dramatically relative to the narrow endpoints well established in higher grade tumors. Continual improvement in understanding of prognostic factors is critical to patient/physician outlook as well as the planning of therapeutic interventions.
Multiple series have shown that age is a useful clinical prognostic marker; younger patients have consistently demonstrated better outcome than older patients (37). In fact, patient age has influenced treatment stratification groups in several prospective trials; 1 practice-defining phase 3 study considered all patients older than 40 as “high risk”, regardless of resection type (04).
Patients presenting with seizures generally have a more favorable prognosis, although it is unclear if lead-time bias is responsible for this delineation, as these patients tend to present earlier in the disease course (07). Alternatively, this may be driven, at least in part, by the incidence of seizures in gliomas with favorable molecular biomarkers such as IDH, which were not accounted for in earlier series.
Tumor size has been described as prognostic of survival as well as histologic transformation. Larger tumors are less amenable to full resection and are more likely to overlay “eloquent” neuroanatomic regions, both of which likely contribute to some of the variation in outcomes (35).
With any progressive low-grade glioma, the possibility of “malignant transformation” should be considered. Although a low-grade glioma by itself generally conveys the invasive phenotype of a malignant tumor, the eventual acquisition of high-grade features marks a critical inflection point. Transformed tumors demonstrate high growth rates, therapeutic resistance, and ultimately lead to clinical deterioration, all characteristic of high-grade gliomas (29). Transformation can be inferred radiographically but is typically confirmed via histopathology after reacquiring tissue via biopsy or reresection of a progressive tumor.
The genetic and histologic changes described and incorporated by the updated WHO classification have had a profound impact on understanding of prognostic subgroups amongst low-grade glioma populations. A 2015 analysis of the Cancer Genome Atlas glioma database found that classifying low-grade gliomas into 3 molecular subgroups was more effective at stratifying outcomes than histology alone (05). The 3 subgroups- IDH-mutation with 1p/19q codeletion, IDH-mutation alone, and IDH-wild type, generated unique survival curves, with median overall survival of roughly 9, 6, and 2 years, respectively. A separate 2015 study further classified gliomas (grade II and III) into 5 distinct prognostic subgroups on the basis of 3 tumor markers: IDH mutation, 1p/19q codeletion, and TERT promotor mutation (10). Notably, each study was able to identify a population of tumors described as “low-grade” on histology but which mimicked the clinical behavior of more aggressive tumors such as glioblastoma.
Health-related quality of life is often overlooked as an outcome and should be considered when evaluating low-grade glioma patients, especially given their increased survivorship burden when compared to high-grade glial tumors. In 1 study, patients consistently scored lower on quality of life measures than healthy age-matched controls and 25% reported “serious” problems with cognitive function (01). Poor seizure control was also a common contributor to poor quality of life index. Downstream treatment effects must also be carefully considered in the context of quality of life endpoints: patients with low-grade glioma who received radiation demonstrated measurable declines in attentional functioning, even when exposed to modest dose levels (08). Quality of life is a key consideration in the design of many therapeutic clinical trials for this patient population.
A previously healthy 45-year-old man presented with a 2-year history of episodes of left facial twitching, preceded by a metallic taste. An MRI demonstrated a nonenhancing lesion in the right insular region. He underwent craniotomy and gross-total resection, which revealed a WHO grade II astrocytoma. After surgery, he was followed with serial imaging alone and remained clinically stable. Eight months later he experienced his first generalized seizure and was found to have new enhancement adjacent to the resection cavity. Stereotactic biopsy was performed, which redemonstrated the initial histologic findings, grade II astrocytoma. Molecular analysis was performed, showing an IDH-1 mutation, 1p/19q intact, and TERT promotor wild type. After a 3-year period of clinical and radiographic stability, an MRI again demonstrated new enhancing components adjacent to resection cavity. Biopsy at this time demonstrated transformation to WHO grade III anaplastic astrocytoma, which was again resected fully. Surgery was followed by 12 cycles of adjuvant 5-day temozolomide, which were well tolerated. After an additional 3-year period of stability, a follow-up MRI showed a mildly enhancing 1 cm satellite lesion in the right temporal lobe, which underwent resection and redemonstrated grade III histology. The patient proceeded with intensity modulated radiation therapy, dosed at 5940 cGy over 6 weeks, given along with concurrent daily temozolomide. Radiation was followed by 6 additional cycles of adjuvant 5-day temozolomide, without further progression. At most recent follow-up, the patient was fully functional with KPS of 90, and had stable imaging.
Low-grade gliomas have been historically described pathologically by their appearance on light microscopy with H&E staining. Oligodendrogliomas appear as moderately cellular and diffusely infiltrating. The terms “diffuse” and “infiltrating” refer to spread of tumor cells into nontumorigenic brain parenchyma, which can be appreciated on microscopy. These tumors are almost always supratentorial, most often in the frontal lobe. Oligodendroglial cells morphologically resemble oligodendrocytes, with rounded nuclei, swollen cytoplasm, and perinuclear haloes (fried-egg appearance) accompanied by delicate branching vessels. Brisk mitotic activity, microvascular proliferation, and spontaneous necrosis should not be present; these are hallmarks of WHO grade III anaplastic oligodendroglioma (23).
Astrocytomas demonstrate mild to moderate cellularity and are similarly infiltrative with difficult to define boundaries on microscopy and MRI. Nuclear atypia is a hallmark, as is their prominent cytoplasmic processes that form a fibrillary stroma. They stain sharply with glial fibrillary acidic protein and vimentin (34). Like oligodendrogliomas, high mitotic index, vascular proliferation, and necrosis are definitive of anaplastic histology.
Oligoastrocytomas display features of both pure tumor types but are considered an “exceptional” diagnosis by the WHO because most glial tumors with mixed histology can be genetically classified as either oligodendroglioma or astrocytoma.
Cellular origin of low-grade gliomas is presently undefined but tumors likely arise from transformed mature/differentiated glial cells, neural stem cells, or glial precursors. Genetic signaling may be a stronger driver than cell of origin, given that oligodendroglioma precursor cells may give rise to either astrocytic or oligodendroglial lineages depending on molecular signature (22).
The 2016 edition of the WHO classification integrated molecular advances into the formal definition of low-grade glioma subtypes. In current practice, presence of IDH mutation defines the class of low-grade glioma. 1p/19q codeletion strongly corresponds to oligodendroglial histology (and is now a definitive requirement for the diagnosis of oligodendroglioma) whereas mutations in TP53 and ATRX are mutually exclusive of 1p/19q loss and are associated with astrocytic histology (05).
When encountered, the population of WHO grade II glial tumors lacking IDH mutation (< 10%) should arouse suspicion and prompt more detailed molecular testing. Most of these tumors will carry molecular imprints of glioblastoma (ie, EGFR amplification), diffuse midline glioma (histone mutation), or occasionally WHO grade I variants such as pilocytic astrocytoma (BRAF fusion) and pleomorphic xanthoastrocytoma (BRAF mutation). IDH-wild type low-grade gliomas with genetic and radiographic markers of glioblastoma should be treated as high grade lesions.
The majority of IDH mutations involve the IDH-1 gene, specifically R132H substitution which is readily detected on immunohistochemistry (46). Because IDH mutation is present in virtually all tumor cells, its presence likely signifies a “driver” mutation of early tumorigenesis (20). Although the specific mechanism of IDH mutation in tumor formation is unknown, its presence confers a well described metabolic disruption – mutated IDH-1 produces a precise enzymatic alteration in Krebs cycle metabolism, leading to accumulation of aberrant byproduct 2-hydroxyglutarate (D2HG). D2HG engenders DNA hypermethylation by inhibiting histone demethylases and TET enzymes. Subsequent dysregulation of DNA methylation dynamics can inhibit normal differentiation processes and confer oncogenicity (45).
The vast majority of astrocytomas occur sporadically. Rarely, they are associated with germline cancer predisposition syndromes such as Li-Fraumeni, Turcot, and Lynch syndromes (30). The only clearly described environmental risk factor is history of exposure to ionizing radiation (32). “Secondary gliomas” are seen in low-grade glioma patients with history of targeted or whole brain radiation therapy, such as those with childhood leukemia or CNS embryonal/germ cell tumors.
Using the 2007 WHO classification criteria, the 2010 to 2014 CBTRUS brain tumor registry data reported that low-grade gliomas (WHO grade II astrocytoma, oligodendroglioma, and oligoastrocytoma) accounted for 16% of all glial tumors. Yearly incidence of all subtypes combined was approximately 1.0 per 100,000 per year. “Pure” astrocytic tumors (0.51 per 100,000 per year) had higher reported incidence than oligodendroglial tumors (0.25 per 100,000). The median age-range of diagnosis was 30 to 40 years, and oligodendrogliomas were more likely than astrocytomas to present over age 40. There was a slight male predominance. All subtypes of low-grade glioma had higher incidence amongst whites compared to black, Hispanic, Asian Pacific Islander, and American India/Alaskan native populations (31).
Moving forward, epidemiologic data should reflect subgroups described in the 2016 WHO classification.
No known preventative measures.
Low-grade gliomas should be distinguished from other primary brain tumors, including higher grade lesions such as glioblastoma and anaplastic astrocytoma and lower grade tumors such as pilocytic astrocytoma, pleomorphic xanthoastrocytoma, and ganglioglioma.
In some circumstances, on MRI, a low-grade glioma can mimic lesions of nonneoplastic etiology such as those from inflammation, infection, stroke, or metabolic disease (16; 26; 14; 40). When glioma is present on such a differential, tissue and histopathology should be obtained to confirm or reject this suspicion.
Low-grade gliomas are generally diagnosed through histology and molecular studies from tissue obtained from biopsy or wider resection. In adults, these tumors rarely present with leptomeningeal involvement; there is no role for lumbar puncture in the absence of specifically localizing findings once the diagnosis is formalized. Routine or surveillance EEGs can be deferred if seizures can be characterized clinically because localization is achieved more precisely through identification of tumor radiographically.
Radiography is essential in the workup and evaluation of low-grade glioma, as imaging helps formulate a differential prior to formal diagnosis of any brain tumor. MRI in particular plays a central role in surgical and radiation planning, and is a main modality used to measure disease response along with clinical history and exam. Low-grade gliomas typically appear with low attenuation on CT, and those with oligodendroglial histology have a propensity to demonstrate calcifications. On MRI, tumors commonly appear centered in the white matter as T1 hypointense and T2 or FLAIR hyperintense. Roughly one third of low-grade gliomas may have T1 postcontrast enhancing components (36).
Advanced imaging modalities, such as MR perfusion, MR SPECT, and PET, have been studied in the context of low-grade glioma and may be contributory when there is a need to differentiate between glioma and nonneoplastic imaging findings (15; 21; 27). At present, there is no clearly defined role for these modalities outside of research settings once a diagnosis is established histologically.
Surgery. Patients with imaging features suggestive of low-grade glioma should have tissue obtained for diagnostic confirmation and molecular analysis, either through stereotactic biopsy or open resection. Although there is no high-level evidence that early tissue obtainment improves outcomes over “watch and wait” observation, in practice it is important to distinguish between low-grade glioma and its mimics, which include other tumor classes and grades as well as nontumorigenic lesions. Early observation could potentially delay diagnosis of an underlying grade III or IV glioma, which require timely and aggressive intervention. In fact, up to 50% of anaplastic gliomas present without enhancement; these may be mistaken on imaging for lower grade tumors (29). There are some lesions, such as those in the brainstem or tectal plate, that preclude safe biopsy and can be reasonably observed in the upfront period (13).
Techniques such as intraoperative MRI, functional MRI, and awake craniotomy are utilized to maximize resection volume while minimizing surgical morbidity. In the presence of mass effect, focal deficits, or growth seen on serial imaging, such maximal safe debulking resection is performed. However, for the population of patients who present incidentally or with seizures controlled by a single antiepileptic drug, it has not been proven that upfront resection leads to favorable outcomes compared to stereotactic biopsy alone.
One of the primary goals of surgery is to establish histologic diagnosis. A wider resection does theoretically improve diagnostic precision – it is not unusual for low-grade gliomas to harbor “islands” of anaplasia that could be missed with biopsy alone (28). An additional goal of surgery is to control seizures – several large retrospective studies have suggested that gross-total resection is a critical factor in achieving seizure freedom (47; 07; 11). There is also evidence that resection may lead to improved survival compared to biopsy alone. A European retrospective in which patients were generally observed until progression following initial surgery demonstrated improved overall survival in the cohort treated with maximal safe resection (19). Several other studies have suggested extent of resection may also influence overall survival (35; 24; 41). In the present era of tumor genetics as a driver of low-grade glioma classification and management, no data yet exists to correlate molecular subtypes with extent of resection and outcome.
Because of the inherent challenges in organizing prospective studies, there is a lack of high-level evidence cultivating population-wide surgical algorithms in low-grade glioma. Although mounting data has shifted the tide towards more aggressive surgical approaches in recent years, it is important to be mindful that these studies are affected by selection bias and generally lack multivariate stratification. Ultimately, all surgical treatment plans are made on an individual case-by-case basis, taking into account age, functional status, location, and patient preference/goals in addition to available data and institutional experience.
Radiation and chemotherapy. Like surgery, therapeutic decisions involving radiation and systemic therapies need to be individualized. Unlike for grade III or IV tumors, it is often acceptable in low-grade glioma to defer such interventions. When given, the dosage, timing, structure, and concurrence of these tools will vary widely between institutions, practitioners, and between individual patients.
Convention has established that initial observation may be appropriate for “low risk” patients; this population is usually defined by age (< 40) and resection quality (gross total resection). Still, a 2008 prospective study from the Radiation Therapy Oncology Group found that adult patients younger than 40 who underwent gross-total resection for low-grade glioma had a greater than 50% risk of progression at 5 years if treated with observation alone (39). The likelihood of recurrence with observation is higher for older patients and for those undergoing less than total resection (06). All patients, regardless of risk profile or treatment exposure, warrant careful clinical and radiographic follow up.
For fractioned radiotherapy, the current standard practice is 50 to 54 Gy to marginated tumor over 6 weeks. A landmark phase 3 European Organization for Research and Treatment of Cancer study published in 2005 randomized low-grade glioma patients to radiotherapy (54 Gy in 1.8 Gy fractions) versus observation (44). Although 5-year progression-free survival was significantly lower in the radiation group (35% vs. 55%), there was no measurable impact on overall survival. Quality of life endpoints were not included in the analysis, so it is unknown whether “progression-free” time was accompanied by deficits in cognition from delayed effects of radiation exposure. This consideration is of particular importance in a patient population that tends to be young, high functioning, and with the potential for extended survival.
An additional finding to come out of the European Organization for Research and Treatment of Cancer study was that early radiotherapy did not induce malignant transformation at a higher rate than observation alone, contrary to commonly held convictions at that time. Roughly two thirds of patients progressed with a higher-grade tumor in both study groups when tissue was available for analysis.
Radiation exposure in low-grade glioma can spur a delayed leukomalacia, best visualized on T2 and FLAIR MRI sequences, which can complicate objective assessments of tumor response (42). Most low-grade lesions are nonenhancing and are monitored primarily with carefully serialized volumetric assessments of T2 signal abnormality. Regions of white matter signal change from treatment effects can overlay and disturb territories of tumorigenic interest, challenging the precision of response assessments by practitioners.
A long-term follow-up of cognitive outcomes in radiation-exposed low-grade glioma patients found that these white-matter hyperintensities, as well as global cortical atrophy, were associated with worse cognitive functioning (08). A clear picture on the degree of radiation-induced cognitive deterioration in low-grade glioma has been elusive, in part because changes in technology, radiotherapy techniques, dose planning, and patient selection make studies difficulty to compare. That said, a clear pattern does emerge from the literature – the incidence and severity of cognitive deficits disproportionately affects older patients as well as those exposed to higher doses and volumes of radiation (43; 03).
Interest in sparing cognitive effects of radiation in part spurred interest in investigating the utility of chemotherapy in low-grade glioma. Several studies in the 1990s demonstrated efficacy of a nitrosurea-based “PCV” (procarbazine, CCNU, vincristine) regimen in anaplastic oligodendroglioma and anaplastic astrocytoma, leading to interest in extending protocols to low-grade glioma (25; 18). Due to toxicity associated with PCV and the more recent introduction of temozolomide as a well-tolerated mainstay in high grade glioma therapy, PCV has fallen off as a first line therapy in most centers.
After some promising phase 2 studies, the European Organization for Research and Treatment of Cancer organized a large phase 3 study that randomized newly diagnosed low-grade glioma patients to radiotherapy or temozolomide. The most recent interim analysis, with a median follow-up of 4 years, found there was no significant difference in progression-free survival between the groups (02). Notably, the subgroup of tumors harboring IDH mutation without 1p/19q codeletion demonstrated inferior outcomes with chemotherapy compared to radiation (progression-free survival 36.0 months versus 55.4 months, p=.004), casting doubt on the utility of upfront standalone chemotherapy in this population.
Without clear evidence-based treatment pathways with radiation and chemotherapy alone, attention shifted to combining modalities, especially given moderate success with radio-chemotherapy in high grade glioma. The most pivotal data has come out of Radiation Therapy Oncology Group 9802, a phase 3 study launched in 1998 that randomized “high-risk” low-grade glioma patients to radiation with or without adjuvant PCV. Although the interim analysis in 2012 did not demonstrate a significant survival difference, a more recent analysis (2016) identified a divergence in the survival tails, with a median overall survival of 13.3 years in the radio-chemotherapy group compared to 7.8 years in the radiation alone group (HR, 0.59; P=0.003) (04). Unfortunately, Radiation Therapy Oncology Group 9802 was designed prior to the present era of molecular stratification in low-grade glioma and reliable data on genetic subgroups is lacking.
Given the current climate favoring temozolomide over PCV in glial tumors, efforts have been made to compare temozolomide with PCV and radiation alone. The CODEL study, initiated in 2009, is recruiting patients to randomize to either radiation with concurrent and adjuvant temozolomide or radiation with adjuvant PCV chemotherapy. Also, in 2009, the Eastern Co-operative Oncology Group (ECOG) commenced a phase 3 study (E3F05) randomizing low-grade patients to radiation alone or radio-chemotherapy with concurrent and adjuvant temozolomide. E3F05 has completed enrollment and has an estimated primary study completion date of December 2021. Due to relatively prolonged survival in the low-grade glioma population, studies can take decades to mature and may lose relevance over time. This is well demonstrated in Radiation Therapy Oncology Group 9802, which was launched in 1998 but wasn’t able to report “positive” primary outcome measures until 2016, at which time PCV had become less frequently utilized and discovery of key molecular markers had fundamentally reorganized the tumor classification scheme.
Radiation Therapy Oncology Group 9802 does provide support for treating high-risk low-grade glioma patients with upfront radio-chemotherapy. Yet the reality of the disease process is that the vast majority of these tumors will resume growing at some point, regardless of therapeutic exposure. Traditionally, tumors treated with either single agent chemotherapy or radiotherapy alone have been treated at progression with the “leftover” agent or modality. With more practitioners combining the modalities in the upfront period, there is a need to develop clearer strategies for combatting recurrent or refractory tumors. Although anti-VEGF agent bevacizumab has demonstrated activity in glioblastoma, this was not replicated in a phase 2 study for low-grade tumors (17).
Targeting of the IDH-1 mutation may be a fruitful path forward because of its ubiquity in tumor cells and likely role as a driver of glioma-genesis. Studies targeting inhibition of IDH-1 mutation enzymes are currently underway and data from the dose expansion cohort of a phase 1 study evaluating IDH-1 inhibitor ivosidenib in progressive IDH-1 mutant low-grade glioma have been presented. For the 24 patients in the expansion arm, the average 6-month tumor growth was 24% prior to treatment and 11% following treatment with ivosidenib. At present, there are additional interests in immunotherapy vaccine approaches targeting IDH-1 antigens.
Rimas V Lukas MD
Dr. Lukas of Northwestern University Feinberg School of Medicine received honorariums from Novocure for speaking engagements, honorariums from Novocure and Merck for advisory board membership, and research support from BMS as principal investigator.See Profile
Nearly 3,000 illustrations, including video clips of neurologic disorders.
Every article is reviewed by our esteemed Editorial Board for accuracy and currency.
Full spectrum of neurology in 1,200 comprehensive articles.
May. 10, 2022
May. 01, 2022
Apr. 07, 2022
Mar. 29, 2022
Mar. 16, 2022
Mar. 11, 2022
Mar. 10, 2022
Mar. 10, 2022