The Effect of Race on the Prognosis of the Glioblastoma Patient: A Brief Review

Abstract

Glioblastoma is the most common primary malignant brain tumor in adults, and despite decades of intensive research regarding its pathophysiology and treatment, the prognosis for glioblastoma patients remains poor. While many studies have analyzed various factors that may influence survival outcomes, the focus of this brief review is to discuss the influence that a patient’s race/ethnicity has on survival. This factor has been investigated in large population-based studies and in smaller institutional analyses, but the prognostic utility of this factor has been inconsistent. Discussion of this topic is therefore warranted to better equip providers to counsel and treat patients with glioblastoma, as well as to identify areas of future research. Although some discrepancies exist, a significant survival benefit is associated with the Asian or Pacific Islander (API) race, whereas white patients have the poorest survival and highest incidence. Hispanic patients tend to fare better than white patients but have worse survival than APIs. Further analysis into the differences in survival among different races may lead to an increased understanding of potential molecular and genetic targets, thus guiding future treatment plans for these patients.

1. Introduction

Glioblastoma (GBM) is the most common primary malignant brain tumor in adults with an average annual age-adjusted incidence rate (AAAIR) of 3.21 per 100,000 population from 2011 to 2015 and an expected 13,310 newly diagnosed cases in 2019.[1] The five-year survival rate for GBM patients is 5.6%, and to this date, patient age and extent of resection (EOR) are two of the strongest prognostic factors.[2, 3, 4] Specifically, younger age at diagnosis and maximal safe tumor resection are associated with significantly improved survival.[5] Other factors portending increased survival include high performance status and O6-methylguanine DNA methyltransferase (MGMT) promoter methylation.[5]

A patient’s race/ethnicity is an additional factor that may affect long-term survival. Incidence of GBM in the United States varies greatly by race/ethnicity. Of note, the AAAIR for white patients is almost twice that of black patients (3.47 versus 1.8 per 100,000 population). Although Hispanic patients have an AAAIR of 2.4 per 100,000 population, white Hispanics specifically have almost double the incidence rate of black Hispanics (2.46 versus 1.29 per 100,000 population).[1] The impact of these factors on patient survival has been investigated in large population-based studies via national databases as well as in smaller retrospective institutional analyses. The demographics from these studies are provided in Table 1. Since the findings of these studies have demonstrated inconsistencies, an exploration into the differences of survival among the different races after surgical resection of an intracranial GBM is warranted.

Table 1.

Recent studies analyzing the association between patient race/ethnicity and survival in glioblastoma patients

		Patient Data Information			Number of Patients by Race & Ethnicity
Yr	Author	Source	Date Range	Age	Total	White	Black	Hispanic	API	AI/AN	O/U
2018	Ostrom	CBTRUS SEER	2000-2014	≥18 yrs	150,631	128,976	8,290	9,217	2,929	525	N/A
2018	Nizamutdinov	S&W BTR	1976-2013	All ages	645	562	38	27	N/A	N/A	18
2018	Huang	SEER	2000-2013	≥20 yrs	30,034	26,878	1693	N/A	N/A	N/A	1463
2018	Bohn	SEER	2010-2014	≥18 yrs	3,473	2,890	205	192	186	Exc	N/A
2017	Xu	SEER	2004-2013	≥18 yrs	24,262	21,642	2446	N/A	1,295	100
2017	Trifiletti	NCDB	2004-2014	≥18 yrs	27,865	23,871	1,535	1,485	532	40
2017	Shah	UM	1995-2006	18-65 yrs	140	47	6	27	N/A	N/A	58
2017	Shabihkhani	SEER, UCLA	2001-20111, 2000-20142	>19 yrs	21,1841 5712	17,2921 4472	11661 62	1,8261 662	N/A1 362	N/A	9001 162
2017	Doung	UCLA	2013-2017	≥18 yrs	170	141	1	14	14	N/A	N/A
2016	Wu	CUMC	1996-2014	All ages	428	313	21	77	17	N/A
2016	Shah	SEER	1999-2010	≥70 yrs	6,039	5586	235	N/A	N/A	N/A	218
2016	Farah	SEER	1995-2012	All ages	33,204	N/A	N/A	N/A	N/A	N/A	N/A
2015	Pan	SEER	2000-2010	>18 yrs	14,675	11,881	773	1,411	623	N/A	N/A
2012	Thumma	SEER	1973-2008	≥20 yrs	34,664	31,820	1,526	N/A	1,117	87	54
2007	Barnholtz-Sloan	SEER	1991-1999	≥66 yrs	1,459	1,418	41	50	25	N/A	N/A

¹SEER cohort.

²UCLA cohort.

Abbreviations: API = Asian or Pacific Islander; AI/AN = American Indian or Alaska Native; O/U: Other/Unknown; CBTRUS = Central Brain Tumor Registry of the United States; SEER = Surveillance, Epidemiology, and End Results; S&W BTR = Scott & White Brain Tumor Registry; NCDB = National Cancer Database; UM = University of Miami, UCLA = University of California Los Angeles; CUMC = Columbia University Medical Center; Exc: Exculded.

2. Methods

A comprehensive literature search was performed utilizing PubMed, Web of Science, and EMBASE databases from January 2000 to January 2019 to identify studies that reported GBM survival outcomes by race/ethnicity. Search terms included the following: glioblastoma and race, glioblastoma and ethnicity, glioblastoma prognostic factors, glioblastoma survival in Caucasians, glioblastoma survival in African Americans, glioblastoma survival in Hispanics, and glioblastoma survival in Asian or Pacific Islanders. Studies that did not report survival by race or ethnicity and studies in non-English languages were excluded.

3. Race & Ethnicity

3.1. Asians and Pacific Islanders appear to have a survival advantage

The largest and most complete analysis of GBM survival by race or ethnicity in the United States was recently performed by Ostrom et al. and included a total of 150,631 cases of GBM using population-based data from the Central Brain Tumor Registry of the United States (CBTRUS) and the Surveillance Epidemiology and End Results (SEER) database.[6] For patients with GBM, the overall 1-year relative survival (RS) after diagnosis was 41.4%, and the 5-year RS was 5.4%. Asians or Pacific Islanders (API’s) were found to have the highest 5-year RS (8.8%), followed by Hispanic whites (7.8%), blacks (6.8%), and non-Hispanic whites (4.8%). Notably, when the survival analysis was adjusted for age and EOR, API race was associated with significantly increased survival. Furthermore, API’s were observed to have the highest 1-year and 5-year survival rates following surgical resection and adjuvant chemoradiation (67.1%, 10.1%), followed by blacks (60.9%, 9.9%), Hispanic whites (60.4%, 9.5%), and lastly, non-Hispanic whites (57.2%, 6.6%).[6]

Several other population-based studies have also shown improved survival in API’s with GBM.[7, 8, 9, 10, 11, 12, 13, 14] A recent retrospective cohort study utilizing the SEER database analyzed a total of 3,473 patients diagnosed with GBM from 2010-2014, and after adjusting for age, gender, health insurance status, primary site, tumor size, and EOR, the study showed that API’s had significantly improved 3-year survival rate when compared to non-Hispanic whites.[7] Four other SEER-based population studies with much larger sample sizes reiterated a survival advantage in API’s.[8, 10, 12, 13] Xu et al. included 24,262 patients from 2004 to 2013 in their study, and multivariate analysis demonstrated that the API race was significantly associated with improved survival compared to white, black, and American Indian races.[8] Shabihkhani et al. studied 21,184 patients from 2001 to 2011 and found that the group containing API’s had significantly higher 1-year and 5-year survival compared to non-Latino whites; however, the API group also included American Indians (AI) and Alaska Natives (AN), which may have influenced the findings.[10] Pan et al. included 14,675 patients from 2000 to 2010 in their study, and multivariate analysis showed that the Asian/other group had significantly better survival than non-Hispanic white patients. Although the Asian/other group had the highest 2-year survival rate (30.7%) of all racial groups, the authors did not state what races were included in the “other” part of the group.[12] An even larger study by Thumma et al. included 31,820 patients spanning from 1973–2008 and showed that API’s had significantly superior 1-year, 2-year, and 5-year survival compared to white patients.[13] Barnholtz-Sloan et al. also showed significantly increased survival in Asians compared to white and black patients, but their study only included patients aged greater than 65 years.[14] Finally, a retrospective study of 27,865 patients from the National Cancer Database (NCDB) was performed by Trifiletti et al., in which multivariate analysis revealed that API’s have significantly improved survival after gross total resection (GTR) compared to all other racial groups except for Hispanics.[9]

Despite the strong evidence for a survival advantage in API’s with GBM, some studies have failed to show a positive association.[15, 16] An institutional analysis by Duong et al. showed no statistical difference in progression-free survival (PFS) or overall survival (OS) in Asian versus non-Asian GBM patients in Los Angeles; however, the small sample size and lack of multivariate analysis to control for potential confounders likely influenced these results.[15] Some studies have evaluated the survival of GBM patients of various races with respect to the mutational status of their tumors.[16, 17, 18] A meta-analysis by Yang et al. showed that Asians with MGMT promoter methylation had no significant difference in PFS or OS compared to Caucasians. However, the meta-analysis included 46 studies involving the Caucasian population and only 6 studies in Asians, which could have masked a potential survival benefit in the Asian population.[16] MGMT promoter methylation status, as well as isocitrate dehydrogenase 1 (IDH-1) status, were analyzed in the institutional study of 428 GBM patients by Wu et al., and although the study found no significant difference in survival among various racial groups, the authors reported that the significance of molecular status on patient survival cannot be adequately assessed due to the incompleteness of molecular screening in their patient sample.[17] A meta-analysis by Dai et al. further examined the association between IDH-1 mutation and risk of mortality in GBM patients, and although the study showed that IDH-1 mutation is associated with significantly decreased risk of mortality in both Europeans and Asians, the authors found that the risk of mortality decreased twice as much in Europeans (65% decreased risk) versus Asians (32% decreased risk) and thereby indicates that the prognostic significance of IDH-1 mutation is less profound in Asian patients.[18] Despite these studies, further assessment of the association between race and molecular features are warranted via institutional and population-based studies, as well as meta-analyses, to better evaluate racial survival differences in GBM patients.

3.2. Conflicting outcomes in patients with Hispanic ethnicity

Survival in patients with Hispanic ethnicity has varied in recent studies. While most studies appear to show a less profound survival benefit compared to API’s, Hispanic patients appear to have improved survival compared to non-Hispanic white patients.[9, 10, 12] The large population-based study by Trifiletti et al. showed that Hispanic patients have significantly improved survival after GTR compared to white patients.[9] Similarly, multivariate analysis in the earlier study by Pan et al. showed that Hispanic patients have significantly better survival compared to non-Hispanic white patients; notably, the 2-year survival rate in the Hispanic group was the second highest (27.1%) among the racial groups, falling just behind the Asian/other group (30.7%).[12]

The most comprehensive study of Latino GBM patients was performed by Shabihkhani et al., in which survival outcomes of several Latino sub-groups were analyzed using the SEER database.[10] Overall, the study found that Latinos had significantly higher 1-year and 5-year survival rates compared to non-Latino whites, and on sub-group analysis, South/Central Americans were found to have significantly improved 1-year and 5-year survival compared to non-Latino whites.[10] Mexican patients were also found to have a significantly higher 5-year survival rate compared to non-Latino whites, but no statistical difference was observed in 1-year survival.[10] Although Latino patients were found to present at a younger age than non-Latino whites, age did not account for the survival difference on multivariate analysis; furthermore, EOR and radiotherapy did not explain the survival difference among the various racial groups on multivariate analysis.[10]

In addition to their SEER analysis, Shabihkhani et al. also examined a smaller cohort of patients from their institution to investigate the significance of race/ethnicity and molecular features, including MGMT methylation and IDH-1 mutation. However, no statistically significant difference in MGMT methylation or IDH-1 mutation was observed between Latinos and non-Latino whites.[10] This echoed the results of the previous institutional study by Wang et al., in which similar rates of MGMT methylation and IDH-1 mutation were observed between Hispanic and Caucasian patients.[19] Of note, no significant difference in 1-year or 5-year survival was found between Latinos and non-Latinos in the smaller institutional cohort analysis by Shabihkhani et al., possibly due to the higher rate of chemoradiation utilization in this group compared to the SEER population.[10]

The SEER-based study by Koshy et al. investigated whether the administration of temozolomide and postoperative radiotherapy translated into improved survival, and on multivariate analysis, neither race nor ethnicity (Hispanic versus non-Hispanic) were found to be risk factors for OS.[20] However, given that the non-Hispanic group included white, black, and Asian patients, it is likely that the more favorable survival of Asian patients masked the poorer survival in the non-Hispanic white patients. Wu et al. performed a smaller institutional analysis of newly diagnosed GBM patients receiving radiotherapy to determine whether outcomes differ in Hispanic patients, but no significant difference in survival was noted among Hispanic, white, black, and Asian patients.[17] Furthermore, subset analysis of patients on temozolomide showed that Hispanic patients had the shortest median survival (411 days), whereas Asian patients had the longest median survival (638 days).[17]

3.3. White patients have the highest incidence and poorest survival

The study by Ostrom et al. showed that non-Hispanic whites have the highest AAAIR (4.71 per 100,000) for GBM and the lowest 5-year RS rate (4.8%) of all racial groups.[6] The lower incidence of GBM in patients of other races and ethnicities could account for part of this survival difference. Disparities in access to healthcare could also contribute to the survival difference. However, persistence of these survival outcomes in patients who received treatment indicate that healthcare access cannot fully account for survival differences. Specifically, non-Hispanic whites have the lowest 1-year and 5-year RS rates after receiving chemoradiation (57.2% and 6.6%, respectively), and in patients undergoing surgery and chemoradiation, all racial or ethnic groups had significantly better survival than non-Hispanic whites.[6].

A small number of studies have shown improved survival in white patients compared to other races.[21, 22] Even though a small institutional retrospective study in Miami found significantly lower OS in Hispanic patients compared to white patients (13 months versus 24.3 months, respectively), the authors posited that the increased prevalence of comorbidities, particularly hypertension and diabetes, in the Hispanic group could explain the survival difference.[21] A population-based study that compared outcomes in elderly GBM patients in the pre-Stupp era versus Stupp era showed that the Caucasian race was an independent predictor of improved survival in patients receiving the Stupp regimen.[22] However, 92.5% of the patients in the study were Caucasian, and the only other racial groups were black, other, and unknown, with no inclusion of distinct Hispanic or API groups. Other than these studies, the majority of institutional and population-based studies have shown that white patients have the poorest survival after GBM diagnosis. [6, 23].

4. Conclusion & Future Research

Both institutional and database studies have shown clear differences in both incidence and survival in GBM patients based on race or ethnicity. The API race has been demonstrated as an independent predictor of improved survival. Moreover, patients of the API racial group appear to have the highest survival after surgical resection and adjuvant chemoradiation. Patients of Hispanic ethnicity have significantly better survival compared to non-Hispanic white patients, who have the poorest survival of all racial groups. However, the lower incidence of GBM in both API patients and Hispanic patients likely accounts for part of the survival advantage.

While a patient’s race/ethnicity is an important factor that should be considered when counselling patients, areas of future research remain that deserve exploration. One important area of study regarding a patient’s race/ethnicity is further investigation into the differences in molecular features, particularly in larger scale studies, which will likely be more feasible over the coming years as most GBM patients now undergo routine screening for MGMT methylation and IDH-1 mutation. However, more detailed investigations of the genetic makeup and tumor microenvironment of GBM may reveal differences in tumorigenesis based on race/ethnicity, which may account for the difference in GBM incidence among the racial/ethnic groups and also provide targets for personalized therapies. A clearer understanding into the molecular alterations and genetic pathways could be the key to improving therapy and ultimately survival.