Introduction

Accounting for 52% of all primary brain tumors, glioblastoma is the most common and most aggressive. Glioblastomas account for 20% of all intracranial tumors. Disparities in risk factors, incidence, treatment, and follow-up have been shown in the literature for several types of cancer. For patients with glioblastoma, the influence of socioeconomic factors, including gender, race, ethnicity, income level, marital status, and occupation, have been explored in several articles. This issue reviews the literature on socioeconomic status (SES) and glioblastoma.

Proposed risk factors for glioblastoma

Although there are no definable risk factors for glioblastoma, several studies have explored putative risk factors. In one article, 80% of patients with newly diagnosed glioblastoma had detectable cytomegalovirus (CMV) DNA in their peripheral blood, whereas seropositive normal donors and other surgical patients did not show detectable virus. Challenging this finding, another study reported that a series of 5 patients with glioblastoma showed no circulating CMV detected either with reverse transcription polymerase chain reaction or blood culture. Several investigators hypothesized that CMV could be a factor in the genesis of glioblastoma if age at infection is taken into account, because the incidence of both glioblastoma and CMV infection are inversely related to SES. CMV infection in early childhood, which is most common in lower socioeconomic groups, may be protective against glioblastoma, whereas CMV infection in later childhood or adulthood may be a risk factor for glioblastoma. If so, glioblastoma occurrence would resemble paralytic polio, in which low SES, poor hygiene, and early infection are protective. This hypothesis has not been supported in the literature.

Some investigators suggest a relationship between glioblastoma and neurocysticercosis. In a case-control study, 6 of 8 patients with neurocysticercosis and a cerebral glioma had calcified parasitic lesions within and around the tumor. The investigators hypothesized that the intense astrocytic gliosis that surrounds calcified cysticerci, together with the suppression of the cellular immune response induced by cysticerci, may contribute to the development of malignant glial cells in patients with neurocysticercosis. However, the relationship between glioblastoma and neurocysticercosis was not statistically significant.

In terms of behavioral risk factors, a study of cigarette smoking, alcohol intake, and risk of glioma in the National Institutes of Health–AARP (American Association for Retired Persons) Diet and Health Study of 477,095 American men and women found that smoking and alcohol consumption did not increase the risk of glioma.

Socioeconomic status and the diagnosis of glioblastoma

The negative impact of SES on many medical conditions is widely understood; patients without financial resources, who lack transportation, and who do not have the means to access timely primary care services often experience delays in diagnosis. Early diagnosis in patients with glioblastoma of differing SES has not explicitly been explored in the literature. However, several articles have examined the incidence of glioblastoma and its relation to SES. For instance, one study found associations between age, sex, Medicaid enrollment, and the incidence of primary malignant brain tumors in Michigan from 1996 to 1997 examining the Michigan Cancer Surveillance Program (1006 cases). Persons enrolled in Medicaid were more likely than nonenrolled persons to develop a malignant brain tumor of any type, including glioblastoma or astrocytoma. Incidence rates for malignant brain tumors in persons enrolled in Medicaid peaked at a younger age.

Another study examined the incidence and epidemiology of glioblastoma in Los Angeles County, California, from 1974 to 1999. Men had a higher overall incidence of glioblastoma compared with women. In this study, non–Latin American white people had the highest incidence rates (2.5 per 100,000) followed by Latin American white people (1.8 per 100,000), and African Americans (1.5 per 100,000). Glioblastoma incidence increased in Los Angeles County after 1989, suggesting that the introduction of MRI may have contributed to an increase in diagnosis. Older age, male gender, higher SES, and non–Latin American white race increased the risk of glioblastoma in this article. Similarly, another study examined 880 patients with a diagnosis of glioblastoma who had surgery between 2006 and 2012; increasing glioblastoma incidence was associated with increasing wages, employment status, lower population density, and greater ownership of cars.

Several other studies have found an association between higher SES and a higher risk of glioblastoma. For instance, one study showed that, relative to persons living in census areas with the lowest SES quintile, the highest SES quintile had a higher rate of glioblastoma. Similar associations were seen in population subgroups defined by age, sex, and race. A study of men aged 20 years and older who died of brain tumors in Washington State between 1969 and 1978 found that increasing SES was associated with a higher risk for all brain tumors, including gliomas and astrocytomas. After adjustment for SES, stationary engineers were found to be at excess risk across all tumor types. Increased numbers of astrocytic tumors were observed for petroleum refinery workers, forestry workers, and cleaning service workers. The strong association between higher SES and glioblastoma risk is unlikely to represent an ascertainment effect because glioblastoma is rapidly progressive and ultimately fatal. An ascertainment effect represents a systematic failure to represent equally all classes of cases or people supposed to be represented in a sample (ie, a sampling bias). In this context, an ascertainment effect would suggest that lower SES patients have not been appropriately included within the sampling, such that it seems that higher SES is correlative as a risk factor for glioblastoma. The investigators of this study do not believe that ascertainment effect explains these findings, and suggest instead that several previously proposed glioma risk factors may be correlated with SES, including atopy and allergy rates.

In terms of brain tumors among children, several investigators reviewed the California Cancer Registry to examine primary central nervous system tumors (PCNST) from 2001 to 2005. Children younger than 5 years old had the highest incidence of malignant PCNST (2.6 per 100,000). Teens aged 15 to 19 years had the highest incidence of benign PCNST (1.8 per 100,000). There was no statistically significant difference in the incidence of malignant PCNST by race/ethnicity in any age group assessed by this study.

Socioeconomic status and the treatment and follow-up of glioblastoma

Studies have revealed that glioblastoma long-term survival was most related to prognostically favorable clinical factors, in particular young age and good initial functional performance score, as well as O(6)-methylguanine-DNA methyltransferase promoter hypermethylation. Socioeconomic, environmental, and occupational factors did not correlate with survival. The influence of site of treatment has been explored in the literature. In one study, survival of patients treated in private hospitals was statistically superior to that of patients treated in public hospitals.

A retrospective study of Michigan Medicaid and Medicare patients with a first primary astrocytoma diagnosis between 1996 and 2000 revealed that controlling for age, income, surgical intervention, comorbidities, gender, and stage, African Americans were less likely to report having received radiation treatment than white people. Patients with dual eligibility in Medicare and Medicaid were less likely to report having received radiation treatment than those with Medicare only. These differences were not seen with chemotherapy. When only those with a glioblastoma were examined, dual eligible patients and African Americans were much less likely to report radiation treatment. These data suggest that disparities in race and insurance status may exist in receiving standard-of-care treatment of astrocytomas. Another study examined 22,777 patients diagnosed with glioblastoma between 1988 and 2007. Overall, 74% received radiation, whereas 26% did not. Factors associated with omission of radiation included older age, lower annual income, African American race, Hispanic race, Asian American race, unmarried status, and subtotal resection/biopsy. The use of radiation was significantly associated with improved overall survival (2-year survival, 14.6% vs 4.2%). Several investigators evaluated surgical intervention and postoperative radiation therapy receipt, and patient data for patients diagnosed with glioblastoma were obtained from the years 2004 to 2008 from a National Cancer Institute database. Younger, married patients in hospital services areas with higher median incomes were significantly more likely to receive both gross total resection and postoperative radiation therapy. The density of radiation oncology–equipped hospitals was also a significant predictor of postoperative radiation therapy receipt. These findings suggest regional variations in of neuro-oncology services and income.

A study of the Surveillance, Epidemiology, and End Results program from 2000 to 2010 of 26,481 patients found that, from lowest to highest SES quintile, median survival ranged from 5 months to 9 months, respectively. The association between SES and survival was not driven solely by poor outcomes in the lowest SES group; the difference in survival between the middle and highest quintiles was also statistically significant. SES remained highly associated with survival in a multivariable model including patient age, sex, race/ethnicity, radiation therapy usage, and surgery type (gross total resection vs other surgery vs no surgery). This study suggests a strong association between higher SES and increased survival after diagnosis of glioblastoma in the United States.

Another retrospective cohort study compared early versus late hospice enrollment of patients with primary malignant brain tumors admitted to the home hospice program of a large urban, not-for-profit home health care agency between 2009 and 2013. Of 160 patients with primary malignant brain tumors followed to death in hospice care, 22.5% were enrolled within 7 days of death. Compared with patients referred to hospice more than 7 days before death, a greater proportion of those with late referral were bed bound at admission, aphasic, unresponsive, or dyspneic. In multivariable analysis, male patients who were receiving Medicaid or charitable care and were without a health care proxy were more likely to enroll in hospice within 1 week of death. Patients with primary malignant brain tumors enrolled late in hospice are severely neurologically debilitated at the time hospice is initiated may not derive optimal benefit from multidisciplinary hospice care. Similarly, a study examining 197 patients with brain tumors found that 53% of patients died at home, 34% at hospice, and 12.5% at the hospital. A positive impact on caregivers for home assistance was recorded in 97% of cases, and 72% of patients had an improvement in their quality of life scores caused by rehabilitation. The decision-making process at the end of life stage is time consuming, but the degree of distress of the family is inversely proportional to the extent of the preparatory period.