Survival rates and patterns of care for patients diagnosed with supratentorial low-grade gliomas

Gliomas are the most common form of primary brain tumor with an incidence rate of approximately 6.5 individuals diagnosed per 100,000 in the United States.1, 2 Although survival from high grade gliomas is generally poor with a median survival ranging from 6 months to a year, individuals diagnosed with low-grade (Grades I and II) gliomas, which account for as many as 25% of all gliomas, may survive for many years3-8 and have a high quality of life during that period. However, a general perception is that all glioma patients do poorly and few data specific to low-grade glioma are readily available to clinicians and patients describing the survival outcomes and treatment patterns for this group.

Several national centralized databases exist in the United States for describing primary brain tumor incidence, survival, and treatment. These include data from the Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute (NCI), the Central Brain Tumor Registry of the United States (CBTRUS), and the National Cancer Data Base with the SEER data being the largest population-based data set.9 Although population-based survival rates for all patients with malignant (and more recently, nonmalignant)1 primary brain tumors are reported annually, more detailed estimates of survival presented by additional tumor and patient characteristics such as grade, age at diagnosis, and race are generally not widely available. The objective of this investigation is to use data from the SEER Registry to provide more accurate estimates of survival for patients diagnosed with low-grade gliomas in the hope that these data may provide helpful information on this topic for both patients and healthcare providers.

MATERIALS AND METHODS

Cases included all patients diagnosed with a World Health Organization (WHO) Grade I or II supratentorial (Topography codes C71.0–71.4, i.e., hemispheric lesions only) glioma (Morphology codes: mixed glioma [International Classification of Disease {ICD}-0 9382], oligodendroglioma [ICD-0 9450] or astrocytoma [ICD-0 9400]) and reported to the SEER Program of the NCI from 1973–2001 (n = 2189).2 To examine a homogenous study population and to reduce the probability of including individuals with metastatic lesions, individuals with more than 1 primary cancer (i.e., a glioma and a cancer of another site) were excluded from these analyses (n = 165). In addition, patients who were diagnosed at death (autopsy only cases) were excluded (n = 15).

In addition to topography and morphology, information on gender, race, age, and year of diagnosis were available as was information regarding whether the patient had received surgical resection (yes/no), radiation therapy (yes/no), and chemotherapy (yes/no) as part of the first course of treatment. Treatment parameters after the first course are not available in these data. Race was defined according to SEER categories of White, Black, and Other because of small sample sizes in the non-Black, non-White categories. Age was used as both a continuous variable (in the proportional hazards model) and a categoric variable in 3 categories: 0–19 years, 20–64 years, and 65 years and older. The primary outcome variable was time to death as measured in years.

Comparison of cases by descriptor variables was done using a chi-square or Fisher exact test for discrete variables and a t test for continuous variables. Estimates of survival probabilities (with 95% confidence intervals [95% CI]) were calculated using Kaplan–Meier product limit methodology and compared using a Wilcoxon log-rank test. Hazard rates were computed using a Cox proportional hazards model.10 All analyses were completed using the SAS statistical software package.11

RESULTS

Descriptive statistics for the sample are presented in Table 1. The majority (69.3%) of cases are classified as astrocytoma with 21.1% and 9.6% classified as oligodendroglioma and mixed glioma, respectively. The reported distributions of these 3 tumor types has changed significantly over time (P < 0.001) with fewer cases being classified as astrocytoma and more being identified as either oligodendroglioma or mixed gliomas (Fig. 1). The majority of tumors were located in the frontal (44%) or temporal (28%) lobes. Fifty-eight percent of the subjects were male. This ratio of males to females (1.4:1) does not vary by age at diagnosis, race, time period, or histology. The mean age at diagnosis is 39.4 years and does not vary by gender or histology but does vary by race (mean age at diagnosis is 40 years for Whites vs. 34 years for Blacks, P = 0.001) and by year of diagnosis with cases from more recent time periods being diagnosed almost 4 years earlier than those from the early 1970s (P = 0.003 for trend). Eighty-nine percent of cases were reported to be White with 5.2% reported as Black. The proportion of Black cases varies significantly (P = 0.002) by age at diagnosis and ranges from 8.7% among patients aged 0–19 years at diagnosis to 2.0% among patients diagnosed at 65 years or older. The proportion of Black cases diagnosed does not vary by year of diagnosis.

With respect to first course of treatment, the largest number (45%) of patients underwent both a surgical procedure as well as radiation therapy. Nineteen percent received only radiation therapy, 28% received only surgery, and 8% received neither. Seventy-six patients were reported to have received chemotherapy as a part of the first course of treatment. The overall proportion of those receiving some form of treatment did not vary by gender, race, or year diagnosed but did differ by age with older patients (> 64 yrs), more likely to receive no therapy. The distribution of treatment options did not vary by gender but did differ by age and race with younger patients and Whites having the highest rates of surgery. Individuals under the age of 65 years were 2.2 times as likely to undergo a surgical procedure than were those 65 years and older (95% CI, 1.6–3.0) while Whites were 1.6 (95% CI, 1.1–2.4) times more likely to undergo a surgical procedure that were Blacks. The use of radiation therapy did not differ by gender or race but was more frequently used in adult patients with patients aged 20 years or older 3.4 times (95% CI, 2.6–4.4) more likely to receive radiation for their lesion than younger patients. Overall, the use of radiation therapy as part of initial treatment has decreased over time (P < 0.01 for trend) with the majority of patients receiving only surgery as the first course of treatment in 1994–2001 (Fig. 2). Treatment differed by location of the lesion, which did not vary by gender or race. Individuals with parietal lobe lesions were more likely to receive radiation therapy (odds ratio [OR], 1.6; 95% CI, 1.2–2.0) and less likely (OR, 0.45; 95% CI, 0.35–0.58) to receive surgical resection than were patients with lesions located elsewhere in the brain (Table 2).

The mean overall survival was 5.9 years (standard deviation = 5.6 yrs) with a median of 4.1 years. Approximately 25% of patients survived for 2 decades. Improved survival was significantly associated with female gender, younger age, White race, year of diagnosis, and histology as well as first course of treatment. Men were 1.18 times (95% CI, 1.04–1.34) more likely than women to have died of their disease. As expected, risk decreased over time from 1974 and increased with age. The age-group 20–64 years was initially dichotomized into 2 groups, 20–39 years and 40–64 years; however, these 2 categories did not differ significantly with respect to survival and were thus combined for presentation of the survival estimates (Fig. 3 and Table 3). Black cases had a death rate 1.4 times (95% CI, 1.1–1.9) that of White cases. Cases diagnosed with mixed tumors and oligodendroglioma had longer survival than did patients diagnosed with astrocytoma. Individuals receiving only surgery as a first course of treatment had the longest survival times, with individuals receiving both surgery and radiation therapy or radiation therapy alone having increasingly lower survival rates, respectively (P < 0.05).

DISCUSSION

The current investigation provides a more detailed presentation of patterns of care and long-term survival outcomes for patients diagnosed with low-grade glioma within the United States than previously available. On average, the low-grade glioma patients in these data survived for approximately 6 years, although variation in survival was quite large with at least 25% of patients surviving for 2 decades. Improved survival was significantly associated with female gender, younger age, White race, and histology as well as by course of first treatment. The finding that women appear to have longer survival times has been previously noted; however, the reason for this difference is not known. It is also well known that increasing age is associated with decreased survival rates.3, 9 It is noteworthy that, when examining the slope of the age-specific survival curves for individuals older than the age of 65 years (Fig. 3), a rapid decline in survival rates is apparent for the first 2–3 years after diagnosis, but this risk levels off in subsequent years, suggesting that a subgroup of patients has a more positive prognosis than the rest. This finding has been reported for all malignant primary brain tumors as a group.4 More detailed information will be required to identify this group with high survival time and is likely to include updated information on pathology as well as genetic markers. Survival rates differed by race with Whites having longer survival times than non-Whites, in particular, Blacks (the only group for which a sample size exists that allows statistical comparison). The reason for this finding is not entirely clear. Whites were older at age of diagnosis and, hence, would be expected to have shorter survival. Although Whites and Blacks were equally likely to receive treatment, the nature of that treatment differed by race with Whites more likely to receive at least 1 surgical procedure, a treatment associated with increased survival. One possible explanation for this finding is that Black patients presented to medical care at a later stage than did Whites. Unfortunately, the database does not include such information and, hence, an examination by stage is not possible. Overall, the proportion of Black patients varied significantly by age; there were few elderly Black patients reported, and additional study of these variations by race are warranted.

It was noted in these data that the distribution of reported histologic subtypes has changed significantly over time with fewer low-grade gliomas being classified as astrocytomas and more being diagnosed as either oligodendrogliomas or mixed gliomas. This change is most likely due to revisions in pathologic classification techniques rather than any real change in pathology.12 This change in classification has been driven in part by the realization that sensitivity to chemotherapy and, hence, survival time is related to pathologic subgroup and molecular characterization, in particular, for patients diagnosed with oligodendroglioma and who have deletions in chromosomes 1p and 19q.13, 14 Despite these revisions, the lack of a universal classification scheme for brain tumor subtypes is an important issue within national databases and was the theme for a recent Brain Tumor Epidemiology Consortium Meeting.3 Although all cases reported to the SEER registries are confirmed by pathology report, these reports are not uniformly rereviewed and, thus, represent opinions of a wide range of pathologists who may have varied levels of experience in the diagnosis of neurologic disorders, a difficult task in the most experienced of hands. In a recent assessment of brain tumor cases from the Connecticut Tumor Registry (a SEER site), the absolute concordance rate for pathology review was 77% when all gliomas were treated as a group but lower when subgroups (i.e., astrocytoma, oligodendroglioma, and mixed gliomas) were considered with mixed gliomas having relatively high discordance rates relative to pathology review.12 Although histology was a significant predictor in these data, it is clear that presentation of survival statistics by histologic subtypes is not yet appropriate given limitations in data accuracy with the national registries. It is also not clear to what extent changes in histologic classification will assist with identification of the 25% of patients who survived for over 2 decades.

It is apparent that the pattern of care for patients with low-grade glioma, at least with respect to the first course of treatment, has changed over time. Patients undergoing surgery in more recent years were less likely to receive concurrent radiation therapy with this surgery than were patients diagnosed in earlier years. Whether this decrease in the use of radiation therapy as an initial therapy reflects a change in the extent of surgical resection, i.e., a greater effort on the part of surgeons to pursue gross total resection rather than a biopsy and radiation, or simply a delay in the timing of radiation until tumor progression is noted (thus potentially delaying the cognitive affects of this therapy) is not known, as detailed information on extent of surgical resection is not available for the majority of patients included in these data. Although no clinical trial data randomizing patients by surgery exist for patients with low-grade glioma, a recent review15 of literature that examines whether extent of surgical resection is a significant prognostic factor in defining survival for patients diagnosed with a low-grade glioma concludes that resection may indeed be a positive prognostic factor for survival, although the evidence is limited15, 16 and a formal metaanalysis difficult given limited statistical information available across studies. Data such as these are likely to influence healthcare providers in selecting surgery as the first course of treatment for this group of patients.

As may be expected, selection of treatment modality was associated with anatomic location affected, with surgery less likely to be used as a first course of treatment in patients with tumors in the parietal lobe, an area of the brain with high functional importance including centers which control an individual's speech and motor capabilities. Of further interest is whether treatment was also a function of laterality, i.e., whether surgery is less likely to be attempted on eloquent areas of the brain such as the left temporal lobe, which generally harbors speech; unfortunately, only limited information is available on laterality in these data. First course of treatment was associated with outcome with individuals undergoing only surgery having the longest survival. Although this finding suggests an association, the question of whether surgical resection provides the greatest benefit to patients with low-grade glioma cannot be unequivocally determined by these data. The answer to this important question5-8, 15-24 will need study within the context of a randomized clinical trial or large population-based study.

In addition to clinical and treatment variables, it is becoming clear that both response to treatment and the associated variation in survival depends upon genotype as well as the interaction between genotype and treatment with individuals with certain genetic polymorphisms having a significantly increased survival time both overall and when treated with certain radiotherapeutic or chemotherapeutic regimes.13, 14, 25, 26 In general, individuals with compromised DNA repair genes appear to have increased survival.26 Hence, the survival estimates reported here are preliminary and do not yet incorporate detailed pathologic, genetic, or environmental exposure data, as this information is not yet available on a large population.

The probabilities presented here indicate that although almost all patients diagnosed with low-grade gliomas eventually succumb to their disease, a high proportion live for several decades. From our clinical experience, most patients remain highly functional with respect to their daily activities during much of the course of their disease. These data suggest that the clinical course of low-grade glioma may be more encouraging than perceived by many and that these patients are good candidates for continued expectant management.