A 31-Year-Old Woman With a Transformed Low-grade Glioma
- Peter C. Warnke, MD, Discussant
- Author Affiliations: Dr Warnke is Visiting Professor of Surgery, Harvard Medical School, and Neurosurgeon in Chief, Division of Neurological Surgery, Beth Israel Deaconess Medical Center, Boston, Massachusetts.
-
Corresponding Author: Peter C. Warnke, MD, Division of Neurological Surgery, Beth Israel Deaconess Medical Center, 110 Francis St, Ste 3B, Boston, MA 02215 (pwarnke@partners.org).
Abstract
Low-grade gliomas in adults have an incidence of 0.8 to 1.2 per 100 000, and their causes are unknown. Despite their histological classification as low-grade, they cannot be cured by any current treatment mode, and no class I evidence exists to guide initial treatment of these tumors. Median survival ranges between 7.5 years and 10 years, with a 5-year survival probability between 55% and 86%. The prognosis depends on age, World Health Organization (WHO) tumor grade, Karnofsky performance score, cytological type (oligodendroglioma vs astrocytoma), and, potentially, the extent of resection. Oligodendrogliomas with loss of heterozygosity on chromosomes 1p and 19q have a distinctly more favorable prognosis and therapeutic response rate. Low-grade tumors progress to high-grade gliomas with aggressive biological behavior at increasing frequency with advancing age. Ms P is a young woman with a previously treated oligodendroglioma, WHO grade II, with loss of heterozygosity on chromosomes 1p and 19q, which at a third resection had transformed into an oligodendroglioma of WHO grade III. She wants to know her current and future therapeutic options.
- KEYWORDS:
- BRAIN NEOPLASMS
- COMMUNICATION
- DIAGNOSIS, DIFFERENTIAL
- EMPATHY
- GLIOMA
- NEOPLASM RECURRENCE, LOCAL
- OLIGODENDROGLIOMA
- QUALITY OF LIFE
- RADIOTHERAPY
- SEIZURES
DR DELBANCO: Ms Q is a 31-year-old health care professional who is married and has children. She is currently not working and has commercial health insurance.
Without history of seizure disorder, Ms Q experienced a generalized tonic-clonic seizure about 30 months prior to this conference. Thereafter, Ms Q's physical examination results were normal, but computed tomography (CT) and magnetic resonance imaging (MRI) scans showed a right temporal mass with surrounding mild edema (Figure 1A), and the lesion was thought to likely represent a primary brain tumor. Ms Q and her physicians concluded that resection would be her best option.
A, Fluid-attenuated inversion recovery (FLAIR) magnetic resonance imaging (MRI) of Ms Q's tumor at diagnosis. The white signal-intense areas in the right temporal lobe correlate with the tumor. Note the lack of a clear demarcated border. B, Left panel, FLAIR MRI of Ms Q's tumor at first recurrence. The tumor has regrown in the dorsal wall of the previous resection cavity, seen as a small dark zone (yellow arrowhead) in front of the FLAIR signal. Note the more circumscribed appearance, suggesting a more delineated tumor growth. Right panel, arterial spin-labeling (ASL) MRI reflecting blood flow at first recurrence of Ms Q's tumor. The dark areas in the region of the tumor (blue arrowhead) reflect reduced blood flow indicative of a low-grade tumor. C, Arterial spin-labeling MRI at second recurrence of Ms Q's tumor. The tumor still appears dark (blue arrowhead), signifying reduced blood flow compared with the surrounding cortex and white matter, compatible with a low-grade glioma.
Prior to resection, a Wada test (intracarotid sodium amobarbital procedure) demonstrated left hemisphere dominance for language. Ms Q underwent subtotal resection for a World Health Organization (WHO) grade II/IV oligodendroglioma because the tumor was hardly distinguishable from the surrounding infiltrated brain tissue and had no sharp border on MRI. Using the staining index for the anti-Ki67 monoclonal antibody MIB-1, the tumor was shown to have a proliferative fraction of 1% to 2%. Chromosome 1p/19q deletions identified in the tumor indicated a favorable prognosis.
Twelve months later, a follow-up MRI showed recurrence of the mass (Figure 1B [left panel]). Ms Q underwent a second, more extensive resection to remove a macroscopic tumor, and the tissue revealed the same pathologic results as before. Early in 2008, 21 months after her second operation, an MRI showed a third recurrence of the mass, and subsequent resection revealed WHO grade III oligodendroglioma. MIB-1 staining showed that the tumor had a proliferative fraction of 50%. Although Ms Q had a somewhat prolonged postoperative recovery, it was complete and she is currently feeling well while taking oral temozolomide chemotherapy.
Ms Q has never smoked and she does not use alcohol or drugs. There is no history of brain tumor in her family. She has a history of Hashimoto thyroiditis, which currently requires no therapy, and underwent uneventful cholecystectomy for cholecystitis in the past. A recent physical examination performed in the brain tumor clinic was unremarkable, revealing no neurological abnormalities. There were no unusual laboratory findings. Her current medications include levetiracetam, 500 mg twice daily, and temozolomide, monthly, per protocol, with a goal of 18 months of therapy.
MS Q: HER VIEW
After the first surgery, the whole tumor wasn't removed. They thought at first that it was, but it wasn’t. There was a small part of the tumor that was left that they watched for a little bit over a year. And then in June 2007, I had to have a second surgery because it started to grow back.
I had a PET [positron emission tomography] scan, and they thought the tumor was gone. But it was not. I do wish that the doctors could tell me why they can't see the cells. I know they are so small, but while you're in the surgery, why can't the doctors look and see which cells are bad and should be removed? So then in January 2008, I had to again face the fact that I was probably going to have to have another surgery.
After the third surgery, the pathology of the tumor has changed, and it's a grade III anaplastic oligodendroglioma. The 2 options I have now are radiation and chemotherapy or a combination of the 2. Watchful waiting is not an option right now because it is a cancerous tumor. I’m very anxious to have another MRI to see if the tumor is completely gone at this point. There have been more than enough times that it wasn't really gone.
After the surgeries, I felt that I wasn't like myself. For one thing, when I got up, it felt like my head was a lot larger than it really was, that it was a lot more swollen. I knew I was talking normally, but I would ask whoever I was talking to, “Do I sound like myself? Do I sound like I’m making sense?” I felt like I wasn't making sense. And I wasn't able to do things in order. For instance, I’d turn on the water faucet, but then I would realize that I needed to turn on the light. I’d turn on the light, but then turn off the light again when I went to turn off the water faucet. It's those things that you take for granted that you’d just be able to do, something as easy as doing that in order, that you're not able to do in order. I actually had to think about what I was doing. So I didn't feel like myself. I felt like somebody else, or like I’d taken a lot of steps backward.
As a clinician, in my care of patients before my illness, I tried to be empathetic, and I tried to spend time. And I thought I was doing a really good job at that. But then after I had the seizure and after the first time going back to work, I realized that I could be empathetic to a greater extent. Now, instead of trying to get people to shorten their stories, I’m trying to really listen to what they have to say. I think the attitude that I’ve seen with surgeons is they want to solve the problem, and they think, “Now we’ve done this. We’ve solved this problem. We’ve done this miracle. Now, you can move on.” And maybe they don't want to feel like they're not able to heal you, or feel like they're actually not able to deal with your problem.
Having a brain tumor has affected my life in every area. It makes you look at things completely differently because you don't know how much time you have. I worry about my family more than myself all the time. Any time I have something new that I have to deal with, I worry about who's going to take care of my kids, and how I am going to be able to take care of my kids, because they're my responsibility.
AT THE CROSSROADS: QUESTIONS FOR DR WARNKE
What are the epidemiology and natural history in adults of glial cell tumors of the brain? What are their clinical presentation and differential diagnosis and how should clinicians assess the various signs and symptoms patients first demonstrate? What are the indications for biopsy vs excision and what are the current treatment modes? What morbidity may accompany perioperative care and ongoing treatment? What does the future hold with respect to treatment of these illnesses? What do you recommend for Ms Q?
DR WARNKE: Ms Q presented for evaluation after a first seizure. Radiological workup showed a mass in the right temporal lobe (Figure 1A). She was then evaluated by a neurologist, who suggested she undergo a lumbar puncture because the differential diagnosis also included temporal encephalitis. The spinal fluid analysis results were normal, and further neuroradiological workup with MRI scanning with fluid-attenuated inversion recovery (FLAIR) T2-weighted and diffusion tensor sequences indicated a low-grade glioma. To lateralize her language function before her first scheduled surgery, Ms Q had a Wada test, consisting of intracarotid injection of sodium amobarbital to evaluate language and memory function in the injected hemisphere. The patient is awake during the injection and undergoes continuous testing. In view of the fact that the patient was right-handed, and the lesion was on the right side, I would not consider this part of a standard diagnostic workup because the likelihood of ipsilateral language localization in a right-handed person is less than 10%, and this can be clarified noninvasively with a functional MRI.1,2 She then underwent navigation-guided resection.
Ms Q experienced a first recurrence in 2007, and the tumor was again resected using navigation-guided microsurgical resection and the same anatomical approach. Her recent third surgery was with a widened craniotomy, again with navigation. Even before the second recurrence, the tumor showed a pattern of hypoperfusion on MRI (arterial spin labeling) (Figure 1, B [right panel] and C), which is not indicative of malignant transformation. Furthermore, on PET with 18fluourodeoxyglucose measuring glucose use, the tumor was always hypometabolic and showed no signs of malignant transformation. The only indication that the phenotype had become more malignant was an elevated choline/creatinine ratio on the last MR spectroscopy before the third resection (Figure 2). The choline/creatinine ratio can serve as an objective marker for removal of the malignant areas, and the ratio was again normalized after the third resection as measured in the tissue adjacent to the resection cavity. At that point, using imaging criteria alone, the surgeon believed that total resection was achieved. This evolution of a low-grade oligodendroglioma to a malignant transformation in 3 years in a patient younger than 40 years is unusual. The fact that Ms Q had 3 surgeries before implementation of adjuvant treatment also seems unusual. The first surgery was believed to be radical, but she experienced a recurrence within 12 months and, given her young age, a second resection was thought to be necessary to delay adjuvant treatment further, in the hope that the second resection could be more radical. When she had a third recurrence, doubts arose as to the tumor's histology; therefore, a new tissue specimen was necessary, combined with another resection. The most likely explanation for this course is that some malignant cells were present at the first surgery but not sampled or revealed in radiological studies. All treatment decisions were made by the tumor board and not an individual neurosurgeon.
The 3 peaks from left to right represent choline, creatinine, and N-acetylaspartate. In normal brain, N-acetylaspartate is the prevalent peak. In tumors of higher grade, the choline peak becomes the largest and the ratio of choline to creatinine is indicative of the grade. Ratios of 2 or higher are associated with more malignant phenotypes in gliomas. Here, the ratio is 2.87. The tumor otherwise had features consistent with a low-grade recurrence; the magnetic resonance spectroscopy was the only indicator of a malignant regrowth.
Epidemiology and Natural History of Low-grade Gliomas in Adults
Low-grade gliomas originate from the glial tissue within the brain and have either predominantly astrocytic or oligodendroglial differentiation.3 Gliomas are categorized using a 4-tier grading system implemented by WHO.4 The different grades refer to the degree of malignant behavior and the prognosis (Table 1). Adult low-grade gliomas are grade II whereas malignant gliomas are grade III or IV. Grade I tumors are benign, often indolent lesions of childhood-like pilocytic astrocytomas and gangliogliomas. Whereas malignant gliomas show rapid clinical and radiographic progression and a short survival (in the range of 1-3 years), low-grade gliomas (WHO grade II) have a prolonged course over many years. They affect young adults and are characterized by slow growth, a high degree of cellular differentiation, and, in the majority of cases, diffuse infiltration of the surrounding brain tissue.3 Adult low-grade gliomas comprise 25% of all adult diffuse astrocytomas. The incidence of low-grade gliomas is 6 per 1 million adjusted standard population, estimated from countries with a standard cancer registry.5,6 Among low-grade gliomas, the histological entities are astrocytoma, WHO grade II (15%-25%); oligodendroglioma, WHO grade II (5%-18%); oligoastrocytoma, WHO grade II (10%-19%); and, with a significantly lower frequency and predominantly growing inside the ventricles, ependymoma, WHO grade II (3%-9%).7 Rare entities include neurocytomas (0.2%-0.5%) and subependymomas.
Table 1. Adult Low-grade Gliomas4
The natural history of WHO grade II astrocytomas, oligodendrogliomas, and oligoastrocytomas is usually characterized by a plateau phase after diagnosis, when no volume increase is found, and then 3 to 5 years of very slow, if any, progression both clinically and radiographically.8,9 Untreated tumors progress over years in an age-dependent manner.10 In patients older than 40 years, tumors show a higher rate of malignant transformation to WHO grade III or grade IV tumors (33% within 5 years, 54% after 10 years, and 67% after 15 years) and concomitantly change their clinical as well as radiographic behavior, with progressive symptoms, sometimes starting as refractory seizures.11 These data are derived from a study with 239 patients treated between 1972 and 1992, with a median follow-up of 10.3 years. Radiographically, tumors start to show areas of contrast enhancement, signifying the anaplastic focus.12 In functional imaging like PET and regional cerebral blood flow measurements, tumors show areas of hypermetabolism, increased blood volume, and increased cerebral blood flow.13,14,15,16
Low-grade gliomas are tumors of young adulthood, with a peak incidence at about age 35 years, close to the age when Ms Q's tumor occurred. There is no influence of sex or race; no causes or risk factors are known except for neurofibromatosis.17 Although malignant gliomas show a male predominance, this has not been established for low-grade gliomas specifically. For unknown reasons, prognosis worsens with age in all brain tumors.18,19
Ms Q's tumor is atypical: it progressed rapidly in less than 3 years and, despite her young age, transformed into a more malignant tumor during this time. However, her tumor is an oligodendroglioma, which carries a better prognosis than a pure astrocytoma and has 10-year survival rates of up to 50%.20,21,22 Within the oligodendrogliomas, she has a particular molecular fingerprint that is characterized by loss of heterozygosity on chromosomes 1p and 19q. In contrast with oligodendrogliomas without associated deletions, Ms Q's tumor has a significantly better prognosis, exhibiting higher sensitivity to radiotherapy and chemotherapy, particularly to temozolomide and to the procarbazine, CCNU (lomustine), and vincristine (PCV) regimen with combined complete and partial response rates between 62% and 92% using MacDonald criteria of volume assessment based on CT or MRI.23,24,25,26
Clinical Presentation
No pathognomonic sign or symptom provides a clinical diagnosis or even leads physicians to suspect a low-grade glioma, but 68% to 92% of patients come to attention with a first-time seizure as adults. More than 60% of patients have no neurological deficit.27 Ms Q had the classical presentation of a first-time adult seizure. Her seizure semiology indicated a lesion in the temporal lobe (ie, a secondary generalized grand mal seizure). She had a short period of confusion with stereotypic movements that evolved into a grand mal seizure. Among young adults presenting with complex seizures, up to 30% harbor low-grade gliomas.28
The differential diagnoses of patients presenting with first-time seizures include idiopathic epilepsy and other nonneoplastic epileptogenic lesions, such as arteriovenous malformations and cavernomas. Depending on the location of the tumor, a variety of focal neurological deficits may be observed, ranging from minute changes in oculomotor function due to small tectal gliomas to significant motor deficits due to tumors invading the internal capsule or the motor cortex.29,30,31 Only a small proportion of patients show signs of increased intracranial pressure. In addition, cerebral vasculitis, multiple sclerosis, and infectious lesions in the central nervous system are in the differential diagnosis.
Any of the aforementioned clinical presentations require an MRI scan to refine the differential diagnosis and the extent of the lesion, and today an MRI scan is almost always indicated. Although a CT scan may rule out larger lesions or those with a higher tumor cell density, based on clinical experience CT scans are less sensitive than MRI scans. In particular, MRI scans are more effective than CT scans for identifying diffusely growing tumors and small tumors in epileptogenic areas (ie, the hippocampus or the structures of the mesial temporal lobe). While electroencephalograms are used to localize the seizure or to analyze interictal phenomena, they are not an acceptable substitute for MRI.
Magnetic resonance imaging may include T1-weighted and T2-weighted imaging, as well as gradient echo, FLAIR imaging, and MR spectroscopy. Low-grade gliomas present as nonenhancing lesions with T2 hyperintensity and, in the case of oligodendrogliomas, often calcifications. The sensitivity of morphological MRI imaging in detecting a low-grade glioma is 80% to 85%, while specificity is only in the range of 60% to 64% compared with the gold standard of histology.32,33,34,35 Magnetic resonance spectroscopy increases sensitivity by differentiating nonenhancing low-grade gliomas from nonenhancing high-grade gliomas, which, in several series, comprise up to 40% of nonenhancing MRI lesions in patients aged 45 or older. In a case series of oligodendroglial tumors, MR spectroscopy was able to differentiate between high-grade and low-grade oligodendrogliomas without regard to their enhancement pattern.36 Ms Q's tumor was diagnosed by MRI, whereas the CT scan showed a hypodensity but not the other features of a low-grade glioma, corroborating the superiority of MRI in the diagnosis of low-grade gliomas.
Positron emission tomography with fluorodeoxyglucose F18 imaging has successfully identified anaplastic foci within a low-grade glioma, as verified by stereotactic biopsy with coregistration.37 Amino acid PET using either tyrosine or methionine and exploiting the active transport mechanism for amino acids at the blood-brain barrier has been successful in diagnosing low-grade gliomas and in differentiating them from high-grade gliomas.38,39 Amino acid PET is also useful to delineate the extent of biological tumor activity, which is usually underestimated by the MRI scan. In this regard, PET serves 2 purposes: by integration into stereotactic imaging, it can guide biopsies toward the biologically most active and prognostically most relevant parts of the tumor, and by delineating the real extent of the tumor, it could also guide radiotherapy planning more efficiently, thereby avoiding underdosing of the tumor periphery.40,41
Indications for Biopsy vs Excision and Current Treatment Modes
Among neurosurgeons and neuro-oncologists, disagreement about the best management of low-grade gliomas abounds. Collectively, there is much to learn. To establish a histological diagnosis, the decision between stereotactic biopsy and primary resection is not based on high-grade evidence but relies on personal experience and beliefs about the natural history of low-grade gliomas.42,43 Tumors that are resectable only with unacceptable morbidity and mortality (ie, tumors in the basal ganglia, brain stem, or highly eloquent areas of the cortex) are an indication for a primary stereotactic biopsy. At the opposite end of the spectrum, large resectable lesions causing midline shift and mass effect should be candidates for primary resection, with the aim of debulking. Epileptogenic tumors that have caused refractory epilepsy can be an indication for primary resection if epilepsy cannot be controlled medically. In this circumstance, resection is primarily for functional reasons rather than for oncological indications.
In the rest of cases, whether tumors should be biopsied, resected, or partially resected is a difficult decision.44,45,46 While microsurgery, with a combination of functional MRI and electrocorticography, has pushed the boundaries of resectability—particularly for tumors in the vicinity of the primary motor cortex or the primary speech area—it is unclear whether this retards progression and improves overall survival.47 Clearly, it has made neurosurgery in these regions both technically feasible and safer.48,49 However, a robust evidence base to support treatment options in individual patients is lacking. A systematic review by Keles et al50 found that options for treatment are not supported by sufficient evidence to make firm recommendations. While numerous articles from single institutions suggest a potential correlation between extent of resection and survival, none of these studies presents class I evidence; ie, data generated by prospective randomized controlled trials with sufficient follow-up and statistical power to examine the scientific hypothesis. In Ms Q's case, her tumor did not exert any mass effect at the time of first presentation, and the indication for a resection was relative. A stereotactic biopsy would have been an alternative. After the first rapid recurrence indicating a faster-growing lesion, a resection or chemotherapy were the alternatives. Resection was chosen to defer chemotherapy and its adverse effects given her young age. At the second recurrence, a wide resection was warranted, given the suspicion that a shift to a malignant phenotype had occurred.
Although the search term low-grade glioma has 3193 citations in PubMed, no Cochrane review of evidence-based treatment options has been performed. For Ms Q and other adults with low-grade gliomas, no systematic analysis of treatment options has been conducted, and the options range from watchful waiting to radiotherapy, surgery, interstitial radiosurgery, and, in predominantly oligodendroglial tumors, chemotherapy. A Medical Research Council trial comparing “wait and see” vs early radiotherapy was never completed because of problems with accrual.51 A German low-grade glioma observational study led to some recommendations, including debulking for lesions with mass effect, radiotherapy for diffuse lesions after documented progression, and interstitial radiosurgery for circumscribed tumors.51
The different treatment results based on best evidence available are summarized in Table 2. No head-to-head comparison studies of treatment options have been conducted.
Table 2. Treatment Results Based on Best Available Evidence
External Beam Radiotherapy. Fractionated external beam radiotherapy has been evaluated in prospective randomized controlled trials generating class I evidence. However, these trials examined either delayed vs early postoperative radiotherapy (European Organization for Research and Treatment of Cancer [EORTC]) or low-dose (45 Gy) vs high-dose (59.4 Gy) radiotherapy.54,55,56 No class I evidence exists as to the efficacy of radiotherapy per se. Thus, the actual effect of radiotherapy on overall survival has not been measured in a randomized controlled trial, since all patients received radiotherapy at some point in time and no radiation-naive control group exists. A new trial (EORTC 22033-26033) is currently under way to compare the efficacy of radiotherapy vs temozolomide chemotherapy after stratification for genetic 1p loss. This ongoing phase 3 study will hopefully shed some light on the efficacy of radiotherapy vs chemotherapy.
The aforementioned studies show that early postoperative radiotherapy, compared with initiating radiotherapy at the time of recurrence, prolongs progression-free, but not overall, survival. In contrast with the same dosing studies in high-grade gliomas, studies comparing different doses of fractionated radiotherapy show no advantage vs higher doses of radiotherapy. Class II and class III evidence in several studies indicate that postoperative radiotherapy, regardless of the timing of its initiation, has some effect on overall survival. Finally, retrospective studies indicate that radiotherapy improves concomitant epilepsy.62,63
The efficacy of radiotherapy depends to a significant degree on targeting endothelial tissue, which in the brain is unique compared with the extracranial vasculature.64,65 Comparing individuals receiving low-dose vs high-dose radiotherapy,66 scores addressing quality of life are significantly lower in the high-dose group.
Surgical Resection. Surgical resection of low-grade gliomas is currently not supported by any class I evidence; the evidence available stems from class II and III studies. The surgical literature can be subdivided into the prevolumetric and the volumetric era. In the prevolumetric era, gross measures regarding the extent of resection were used for biopsies, subtotal resection, and gross total resection. In contrast, MRI volumetry, both before and after resection, actually quantifies the extent of resection. The prevolumetric era produced inconsistent data, supporting both “the more, the better” and “less is more.” Several studies indicate that patients with resection, compared with those with biopsy, have significantly longer overall survival.67 Others show no effect of resection on survival.46,68 These studies are retrospective, and most lack multivariate analysis of independent prognostic variables, such as age, Karnofsky score (the National Cancer Institute's standard score for tumor patients), WHO subtype, and presence or absence of contrast enhancement on MRI. In an extensive systematic review published in 2001, Keles et al came to a disappointing conclusion: “Currently, for patients with low-grade glial tumors located in the cerebral hemisphere, the only management standard based on high-quality evidence is tissue diagnosis.”50
More recently, quantitative MRI volumetry from defined MRI sequences has provided some evidence that aggressive resection results in improved survival.42,52,53,69 Unfortunately, these studies are again limited by several factors. First, as retrospective, single-institution studies, they are at high risk of patient selection bias. Second, if a small percentage of the tumor is left behind, as in subtotal resection, the survival advantage is lost. Data presented so far show no continuous correlation between the extent of resection and survival; only maximal or gross total resections affect survival. Given that these tumors are invasive by definition, a true, complete “resection” is impossible, raising doubts as to the “threshold” effect of resection. Thus, if small areas of tumor are left behind that just escape MRI detection, as may have been the case for Ms Q, residual tumor should be continuously correlated with survival, rather than there being a threshold at which point the correlation no longer exists. Third, nonprospective, population-based studies of much larger patient cohorts question the effect or extent of resection on survival in patients with low-grade gliomas and raise again the point of selection bias in single-institution series.46
In summary, although surgery is the treatment of choice for space-occupying lesions that cause a neurological deficit, exert mass effect (measured as midline shift), or cause refractory epilepsy, no class I evidence supports a general recommendation that a patient like Ms Q undergo surgery.
Interstitial Stereotactic Radiosurgery. For circumscribed, small gliomas with a maximum diameter of 3 to 3.5 cm, several European groups have used stereotactic, interstitial radiosurgery in large patient cohorts with extensive long-term follow-up.11,65,70,71,72 This technique involves the permanent or temporary implantation of radioactive small seeds into the tumor center, thus using low-dose-rate continuous radiation to the tumor, with a very high dose in the center and a steep dose drop-off to the periphery. Because of its highly localized character, this treatment strategy can apply only to demarcated, nondiffuse tumors that tend to proliferate very slowly. Since the likelihood of damaging a tumor cell in its vulnerable replication cycle is exponentially higher, such continuous hyperfractionated radiation should have a much higher “hit rate.” In addition, DNA repair mechanisms are less effective, resulting in a higher tumor cell kill rate. Several groups have studied the radiobiology of interstitial radiosurgery, and the 5- and 10-year median survival rates in both adult and pediatric low-grade gliomas rival those of surgery and radiotherapy.11,62,71,73,74 Again, like open surgery, interstitial stereotactic radiosurgery is reported by single institutions, thereby increasing the likelihood of selection bias.
Chemotherapy. Chemotherapy, used extensively in pediatric low-grade gliomas, has also been used in adult low-grade gliomas for patients who were either inoperable or refused radiotherapy. In purely astrocytic gliomas, reports on response rates are anecdotal. For oligodendroglioma, there is evidence for chemosensitivity to alkylating agents, such as temozolomide or the PCV regimen.24,57,75,76 In this subgroup of tumors, chemosensitivity appears to correlate with the molecular fingerprint, (ie, the loss of heterozygosity on chromosomes 1p and 19q), and Ms Q has such a fingerprint. The data derive from nonprospective, nonrandomized, uncontrolled studies using objective response criteria from MRI and molecular stratification. Current consensus is that chemotherapy should be used before radiotherapy in younger patients.76,77,78 The good news for Ms Q is that the effect of chemotherapy can go beyond an improved MRI result. It translates into significantly prolonged overall and progression-free survival in patients whose imaging results improve with chemotherapy.
Perioperative Care and Morbidity Due to Treatment
For patients undergoing either biopsy resection or stereotactic interstitial radiosurgery for low-grade glioma, perioperative care varies. Stereotactic biopsy requires local anesthesia, with intraoperative imaging and postoperative CT scans to rule out hemorrhage. Following these procedures, patients generally spend only 1 night in the hospital. For craniotomies performed while the patient is awake, the patient is generally given anesthesia during the craniotomy and surgical opening. For complex resections, patients undergo several hours of general anesthesia and must therefore be stable from a cardiac and pulmonary standpoint. After surgery, such patients require monitoring of cerebral function and usually stay overnight in an intensive care unit. For patients undergoing stereotactic interstitial radiosurgery, perioperative care is nearly identical to that of a stereotactic biopsy. During radiosurgery, radiation monitoring is performed during implantation of the radioactive source, as well as after wound closure, to document and certify that the amount of radiation emitted by the patient reaches legally acceptable levels before the patient can be discharged.
Morbidity from the treatment options differs widely. With the use of sophisticated electrophysiology and navigation systems using functional MRI, which allows the surgeon to visualize brain function on a morphological MRI image, as well as the incorporation of intraoperative MRI and CT, surgical morbidity has fallen to 4% to 6% in modern series, and mortality is usually less than 1%.52,79,80 Stereotactic biopsies carry a morbidity of 2% and a mortality of less than 0.5% if carried out with MRI guidance and image fusion.81 Finally, as might be expected, surgical morbidity also depends heavily on the location of the tumor. Ms Q's reports of changed self-awareness and time perception correlate well with the site of her lesion in the right lateral temporal lobe and are almost classic. Her perception of the neurosurgical mind-set being such that “a problem has been solved” although the problem cannot be solved surgically is a wonderful description of the neurosurgical deformation professionelle and describes the need for neurosurgeons dealing with tumors to address their patients' concerns and goals as part of the treatment approach and follow-up.
Adverse effects from radiotherapy are classified as acute, early delayed, and late delayed. Morbidity is related to total dose, fraction size, and volume of irradiated brain. The most important morbidity is that of symptomatic radionecrosis, wherein brain tissue undergoes necrosis that can be progressive and is accompanied by perifocal edema. Such patients require long-term corticosteroid therapy and, at times, resection of the necrotic mass. The 2-year actuarial incidence is 2.5% after 50.4 Gy and 5% after application of 64.8 Gy.55 Late radiation effects center on leukoencephalopathy and concomitant cognitive decline.82,83 Interstitial radiosurgery is associated with a permanent morbidity of less than 3%, which is dose- and dose rate–dependent.84
Chemotherapy-associated morbidity is drug-dependent, classified according to National Cancer Institute criteria, and ranges between 3% and 25%. Permanent morbidity from the most commonly used drugs (ie, temozolomide or PCV) is in the range of 4% to 8%.85,86 For women of childbearing age, fertility is significantly influenced by these drugs and needs to be taken into consideration because reported infertility rates after PCV chemotherapy are about 50%.87
The Future of Low-grade Glioma Management
Although low-grade gliomas appear histologically benign, their behavior is malignant because no curative treatment exists. I believe that further improvement in outcomes from surgery, radiotherapy, or classic cytotoxic chemotherapy is unlikely. The relative resistance to standard oncological treatment is founded in the unique biology and physiology of low-grade gliomas. In the beginning, these tumors grow slowly and with low proliferative indexes. Histologically, there are no anaplastic features; these gliomas have no border and infiltrate the surrounding brain at a very early stage. They grow within highly functional tissue and have an intact blood-brain barrier. Before they transform into malignant phenotypes, their blood flow and blood volume is low. Thus, they are fundamentally different from most other tumors in the body. If the interspersed normal brain is to be spared, new treatments that target the specific tumor and its biology are needed. Ideally, such an agent both would target tumor cells in the infiltrated brain specifically by binding to a receptor selectively expressed on tumor cells and would not be toxic to normal brain. It would either cross the blood-brain barrier easily or be delivered directly to the extracellular space of the tumor.
Such an agent could either cause cell death or induce redifferentiation of tumor cells and halt the malignant transformation.88 Used so far in malignant gliomas, one approach is to use convection-enhanced delivery of conjugated immunotoxins that bind to either the transferrin receptor or the interleukin 12 receptor, both expressed only by tumor cells in the brain.89,90 Trials in malignant gliomas have not been promising because of several factors, including tumor physiology, with enlarged extracellular spaces and pressure gradients within the tumor. These do not pertain to low-grade gliomas, making them an interesting target for convection. Another intriguing approach is to use stem cells, which have enormous tropism for tumor cells scattered within the brain.91 Finally, in some gliomas, truly interstitial radiosurgery could be performed by attaching radionuclides with a short-range energy to molecules targeting tumor cells via receptor binding.92 Ideally, alpha emitters would be attached, as they irradiate on almost a single-cell level. In summary, as technology advances, I anticipate that glioma treatment will result in more than palliation and will be performed more and more frequently as “molecular neurosurgery.”
RECOMMENDATIONS FOR MS Q
Ms Q has undergone gross total resection of a malignant oligodendroglioma after 2 previous resections for a low-grade glioma in the same location. Without adjuvant treatment, her progression-free interval will likely be short. Adjuvant options are chemotherapy and radiotherapy either alone or in combination. Although radiotherapy for malignant oligodendrogliomas has been the standard for decades and is efficacious in virtually all malignant gliomas, it is associated with a risk of significant irreversible adverse effects, such as radionecrosis and leukoencephalopathy.
At this point, Ms Q's tumor has undergone malignant transformation. However, because her tumor cells have shown loss of heterozygosity on chromosomes 1p and 19q, her prognosis remains favorable and her likelihood of response to chemotherapy is between 62% and 92% since she has a recurrent but chemotherapy-naive tumor.93 In addition, she exhibited hypometabolism on PET with F18, and had a negative thallium 201 scan result. Both are independent prognostic variables associated with response to chemotherapy; hypermetabolic tumors have a median progression-free survival of 22 months vs a median not reached in the hypometabolic group with a median follow-up of 25 months.93,94
In view of Ms Q's age and her favorable prognosis, I recommend deferring radiotherapy and its possible associated adverse effects as long as possible, since an effective, less toxic treatment is available that will likely result in long-term control. I would reserve radiotherapy for a possible point when the tumor becomes chemoresistant or the bone marrow reserve capacity is exhausted. Unfortunately, little can be done for her temporal lobe–related symptoms.
Financial Disclosures: Dr Warnke reports that he has received grants through the Royal College of Surgeons, British Medical Journal, Direct Therapeutics, and ABC Foundation and that he serves as an associate editor for the Journal of Neurology, Neurosurgery, and Psychiatry.
Additional Contributions: We thank the patient for sharing her story and providing permission to publish it.
This conference took place at the Surgical Grand Rounds at Beth Israel Deaconess Medical Center, Boston, Massachusetts, on May 21, 2008.
Clinical Crossroads at Beth Israel Deaconess Medical Center is produced and edited by Risa B. Burns, MD, series editor; Tom Delbanco, MD, Howard Libman, MD, Eileen E. Reynolds, MD, Amy N. Ship, MD, and Anjala V. Tess, MD.
Clinical Crossroads Section Editor: Margaret A. Winker, MD, Deputy Editor.
REFERENCES
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
- 26.
- 27.
- 28.
- 29.
- 30.
- 31.
- 32.
- 33.
- 34.
- 35.
- 36.
- 37.
- 38.
- 39.
- 40.
- 41.
- 42.
- 43.
- 44.
- 45.
- 46.
- 47.
- 48.
- 49.
- 50.
- 51.
- 52.
- 53.
- 54.
- 55.
- 56.
- 57.
- 58.
- 59.
- 60.
- 61.
- 62.
- 63.
- 64.
- 65.
- 66.
- 67.
- 68.
- 69.
- 70.
- 71.
- 72.
- 73.
- 74.
- 75.
- 76.
- 77.
- 78.
- 79.
- 80.
- 81.
- 82.
- 83.
- 84.
- 85.
- 86.
- 87.
- 88.
- 89.
- 90.
- 91.
- 92.
- 93.
- 94.
