Trends in the Management Paradigms of Intracranial Meningioma

Abstract

Objective  Intracranial meningiomas are the most common primary brain tumor. Treatment paradigms have evolved over time. There are limited number of population-based studies that examine this modern evolution. Here, we describe the trends of management of intracranial meningiomas using a national database.

Methods  The data were obtained from the National Cancer Database for the years 2004 to 2015, the collected variables included: patients' age, gender, insurance type, income, comorbidity score, the tumor size and grade, and treatment modality (observation, surgery, radiotherapy, or combination therapy). We performed statistical analyses to detect association between unique variables and outcomes. In addition, we performed mortality analyses for various treatment modalities.

Results  A total of 199,096 patients with a diagnosis of intracranial meningioma were included, the majority of patients were white females, mean age of 61 years, and half of the tumors were ≤ 3 cm. Observation was the most commonly used management modality followed by surgical resection, radiotherapy, and combination therapy. For the entire time period, there was an increased use of observation as a primary management method. Predictors of mortality included increased age, larger tumor size, higher tumor grade, treatment at a community hospital, and higher comorbidity scores.

Conclusion  Population-based studies of intracranial meningiomas are uncommon; our study is one of the few reports that examine the changes in the modern management paradigms of meningioma in the United States over time. Additionally, we shed light on the factors that affected survival of patients with this condition.

Introduction

Intracranial meningiomas are extra-axial brain tumors that were first described in 1614 by Felix Plater in an autopsy report. However, it was Harvey Cushing who first coined the term meningioma in 1922. ^{1 2} Meningiomas are the most common primary intracranial tumors. ^{3 4 5 6} The clinical management of these tumors varies considerably by region, treating physician, patients' age, and tumor size and location. Currently available treatment options include surgical resection, radiation, observation, or a combination therapy. ⁷ There is no Food and Drug Administration-approved chemotherapy currently accepted as part of its management. ⁸ Patient-related factors such as age, functional status, and symptomatology, and tumor-related factors such as location, size, and grade determine the treatment paradigm used. ⁷ Treatment paradigms differ between individual institutions and physicians. ^{3 9}

The National Cancer Database (NCDB) of the American College of Surgeons and American Cancer Society is a clinical oncology database. ¹⁰ The NCDB can be used to analyze the changes of the treatment paradigms over time across a large population. Also, it can be used to assess the variability in treatment methods based on the facility type (academic institution vs. community hospital), region, and insurance type. Here, we describe the findings of our analyses of this database.

Materials and Methods

Data Source

The data for this study were obtained from the NCDB, a database sourced from hospital registry data that are collected in more than 1,500 Commission on Cancer (CoC)-accredited facilities. The NCDB data represent more than 70% of newly diagnosed cancer cases nationwide and more than 34 million historical records. ¹⁰ We identified patients with a diagnosis of intracranial meningioma based on the International Classification of Diseases for Oncology (ICD-O3) using the codes 9530 to 9539. We excluded patients <18 years of age, with bilateral tumors, who received neoadjuvant, intraoperative radiotherapy, or unconventional radiotherapy modalities, and with unknown surgical or radiation treatment status. The study was limited to the period between 2004 and 2015.

Variables

We classified patients according to the treatment type into observation, surgery, radiotherapy, and combination therapy (radiotherapy within 3 months of surgery). Of note, the modalities of radiotherapy used included external beam radiation and stereotactic radiosurgery. Patient-related variables included age, race, gender, Karnofsky score (KPS), comorbidity score (Charlson–Deyo score [CDS]), facility type (academic, community, or network), insurance type (private, government, or none), and urban or rural setting. Tumor-related variables included size, grade, and year of diagnosis. Tumor size was classified into ≤ 3 or > 3 cm. Tumor grade was classified into World Health Organization (WHO) grade 1 benign, WHO grade 2 atypical, and WHO grade 3 malignant. CDS is a comorbid condition scoring system. Its value is a weighted score derived from the sum of the scores for each of the comorbid conditions listed in the Charlson comorbidity score mapping table ( Table 1 ). ¹¹ Academic program is a facility that participates in postgraduate medical education. The facility should add more than 500 newly diagnosed cancer cases each year. Also, it should offer a full range of diagnostic and treatment services. Also, the facility should take part in cancer-related clinical research. Community program is a facility that adds more than 100 but less than 500 newly diagnosed cancer cases each year and should provide a full range of diagnostic and treatment services. Also, it should take part in cancer-related clinical research. Network facility is an organization that owns, operates, or is part of an enterprise with multiple facilities providing integrated cancer care and offers comprehensive services. It should include a minimum of one hospital that has a CoC-accredited cancer program. The facility should take part in cancer-related clinical research either by enrolling patients in cancer-related clinical trials or by referring patients for enrollment at another facility.

Table 1. Charlson comorbidity score mapping table.

Score	Clinical condition
0	None of the conditions below
1	Myocardial infarction Congestive heart failure Peripheral vascular disease Cerebrovascular disease Dementia Chronic pulmonary disease Rheumatologic disease Peptic ulcer disease Mild liver disease Diabetes
2	Diabetes with chronic complications Hemiplegia or paraplegia Renal disease
≥3	Moderate or severe liver disease Acquired immune deficiency syndrome

Statistical Analysis

Statistical analyses were conducted using MATLAB version 9.3 (The MathWorks, Inc., Natick, Massachusetts, United States). Descriptive statistics were reported for each variable. The univariate association with treatment group (observation vs. radiotherapy vs. surgery vs. combination therapy) was assessed using chi-square test for categorical covariates and analysis of variance for numeric covariates. Multinomial logistic regression modeling was used to adjust for potential confounders in identifying predictors of treatment choice. Clinically relevant factors associated with a p -value < 0.10 in univariate analysis were included in the multivariable analysis. To assess the factors significantly related to mortality, a stepwise Cox hazard proportional regression model with predictors was added in a stepwise fashion and retained based on the change in the model deviance. All tests were two sided, and a p -value < 0.05 was considered statistically significant.

Results

Patient Population and Demographics

A total of 199,096 patients with a diagnosis of intracranial meningioma were identified in the NCDB between 2004 and 2015 ( Table 2 ). Of these patients, 73.3% ( N  = 146,205) were females and 76.4% ( N  = 152,159) were white. The mean patient age was 61 years (range: 18–100). Of the entire cohort, 39.3% ( N  = 78,311) had private insurance, 55.2% ( N  = 110,030) had government type insurance, and 3.2% ( N  = 7,796) had no insurance. The median Karnofsky performance score was 75 (normal range: 0–100). ¹² Of all patients, 73.4, 17.9, 6, and 3.1% had CDSs of 0, 1, 2, and ≥3, respectively. The median follow-up was 47 (range: 0–156) months for the observation group, 60 (range: 0–157) months for the radiotherapy group, 59 (range: 0–157) months for the surgery group, and 55 (range: 0–155) months for the combined therapy group.

Table 2. Patient and tumor characteristics.

	Total population	199,096
		Total number	%
Race	White	152,159	76.4
	Nonwhite	45,431	22.8
Gender	Male	54,084	27.1
	Female	146,205	73.4
Facility type	Research	75,699	38
	Community	86,553	43.4
	Network	24,872	12.4
Insurance status	Private	78,311	39.3
	Government	110,030	55.2
	None	7,796	3.9
Urban/rural	Metropolitan	166,230	83.4
	Nonmetropolitan	28,160	14.1
Charlson–Deyo score	0	146,296	73.4
	1	35,633	17.9
	2	12,021	6
	3+	6,339	3.1
Size	≤3 cm	99,681	50
	>3 cm	100,608	50.5

Tumor Size and Grade

Fifty per cent of tumors were ≤ 3 cm in size. For the surgical group, the majority of tumors (94.9%) were WHO grade 1, and only 4.4 and 1.2% were WHO grades 2 and 3, respectively ( Table 2 ).

Practice Patterns

Of the entire cohort, 83.4% were from a metropolitan area; 38% of patients were treated at an academic institution, and 43.4 and 12.4% were treated at a community and network facility, respectively. Of all patients, 51.3% were managed with observation, 5.9% with radiotherapy, 39% with surgery, and 3.5% with combination therapy ( Fig. 1 ). For tumors ≤ 3 cm in size, the majority were managed with observation 73.7%, while for tumors > 3 cm in size, the majority were treated with surgery 60.1% with surgery ( Table 3 ). For patients treated at an academic institution, surgery was the primary treatment modality (43.2%, N  = 32,767), and for patients treated at a community hospital, observation was the primary treatment modality (62.3%, N  = 54.09) ( Table 3 ). The patient comorbidity scores also affected the choice of primary treatment modality, patients with high CDS more likely to be managed with observation (CDS [2] 60.3%, CDS [≥3] 65.8%) and less likely to receive surgery (CDS [2] 35.2%, CDS [≥3] 30.5%) ( Table 3 ) ( Fig. 2 ).

Fig. 1.

A bar chart shows the percentages of the use of different treatment modalities for the entire meningioma patient population.

Table 3. Frequency of use of different treatment modalities according to facility type, patient, and tumor characteristics.

Variable		Total	Observation, %	Radiotherapy, %	Surgery, %	Combination therapy, %
Facility	Research	75,699	43.2	6.8	46	3.8
	Community	86,553	62.3	4.2	31.2	2.1
Tumor size	≤3 cm	99,681	73.7	6.7	18.5	0.9
	>3 cm	100,608	29.8	4.5	60.1	5.4
Charlson–Deyo score	0	146,296	49.8	6.7	39.9	3.4
	1	35,633	53.7	3.2	40.1	2.8
	2	12,021	60.3	1.8	35.2	2.5
	3+	6,339	65.8	1.7	30.5	1.9

Fig. 2.

A stacked bar chart shows the effect of patient comorbidity score on the use of different treatment modalities.

Over time, there was an increased use of observation (39.2% in 2004 and 56.3% in 2015, p -value <0.01), decreased use of radiotherapy (6.7% in 2004 and 4.5% in 2015, p -value <0.01), and decreased use of surgery (51.1% in 2004 and 35.6% in 2015, p -value = 0.01) ( Tables 4 and 5 ). The total number of meningiomas treated increased between 2004 and 2015 (19,636 vs. 11,039) ( Fig. 3) .

Table 4. The frequency of use of different treatment modalities according to different variables.

Variable	Category	Observation ( N  = 102,301, 51.38%)		Radiotherapy ( N  = 11,922, 5.99%)		Surgery ( N  = 77,763, 39.06%)		Combination therapy ( N  = 7,110, 3.57%)		p -Value
		Number	%	Number	%	Number	%	Number	%
Race	White	79,980	78.1	8,799	79	58,697	75.4	4,683	73.2	<0.00001
	Nonwhite	22,321	21.8	2,332	20.9	19,066	24.5	1,712	26.7
Gender	Male	25,221	24.3	2,799	24.6	23,644	29.9	2,420	37.4	<0.00001
	Female	78,309	75.6	8,550	75.3	55,301	70	4,045	62.5
Facility type	Research	32,767	32.8	5,213	48.8	34,825	49	2,894	51.1	<0.00001
	Community	53,968	54	3,656	34.2	27,062	38.1	1,867	32.9
	Network	13,042	13	1,803	16.8	9,128	12.8	899	15.8
Insurance status	Private	28,561	28.1	5,318	47.7	40,888	52.8	3,544	55.6	<0.00001
	Government	69,056	68.1	5,588	50.1	32,866	42.5	2,520	39.5
	None	3,686	3.6	229	2	3,576	4.6	305	4.7
Urban/rural	Metropolitan	87,793	87.3	9,046	83	64,095	83.5	5,296	84.9	<0.00001
	Nonmetropolitan	12,766	12.7	1,847	16.9	12,608	16.4	939	15
Charlson–Deyo score	0	72,955	70.4	9,850	86.7	58,462	74	5,029	77.7	<0.00001
	1	19,146	18.4	1,167	10.2	14,305	18.1	1,015	15.7
	2	7,258	7	224	1.9	4,239	5.3	300	4.6
	3+	4,171	4	108	0.9	1,939	2.4	121	1.8
Size	≤3 cm	73,493	70.9	6,737	59.3	18,476	23.4	975	15	<0.00001
	>3 cm	30,037	29	4,612	40.6	60,469	76.6	5,490	84.9
Age	Years	69.21	14.8	62.01	14.1	57.99	13.9	56.02	13.6	<0.00001
KPS	N/A	63.89	39.8	83.02	22.8	72.47	34.9	78.92	23.3	<0.00001

Abbreviations: KPS, ; N/A, not available.

Table 5. Trend of use of treatment modality over time.

Year	Observation		Radiotherapy		Surgery		Combination therapy
	Number	%	Number	%	Number	%	Number	%
2004	4,327	39.2	747	6.7	5,645	51.1	320	2.9
2005	5,452	43.4	853	6.8	5,862	46.7	369	2.9
2006	6,283	45.6	966	7	6,100	44.3	422	3
2007	7,057	48	960	6.5	6,266	42.6	403	2.7
2008	8,113	49.9	981	6	6,672	41.1	467	2.8
2009	9,272	52	1,093	6.1	6,893	38.7	549	3
2010	9,437	53.4	1,026	5.8	6,557	37.1	639	3.6
2011	10,045	54.6	941	5.1	6,829	37.1	559	3
2012	10,400	55.1	981	5.2	6,864	36.3	626	3.3
2013	10,799	54.9	971	4.9	7,165	36.4	704	3.5
2014	11,276	56.2	938	4.6	7,086	35.3	738	3.6
2015	11,069	56.3	892	4.5	7,006	35.6	669	3.4

Fig. 3.

A marked line chart shows the increase in the diagnosis of meningioma between 2004 and 2015.

Overall Survival

Our survival analysis showed a median overall survival of (140.3) months. Also, we found that patients treated at community hospitals had increased mortality in comparison to academic institutions (hazard ratio [HR]: 1.18, 95% confidence interval [CI]: 1.02–1.36, p -value <0.05). Older age was associated with a higher mortality for all groups (HR: 1.06, 95% CI: 1.05–1.06, p -value <0.05). Female gender was associated with an increase in survival (HR: 0.72, 95% CI: 0.63–0.83, p -value <0.05). A tumor size that is more than 3 cm had increased mortality (HR: 1.45, 95% CI: 1.15–1.83, p -value <0.05) in comparison to tumors equal or less than 3 cm. A higher tumor grade was associated with an increased mortality (WHO grade 2, HR: 1.31, 95% CI: 1.12–1.54, p -value <0.05) (WHO grade 3, HR: 3.31, 95% CI: 2.81–3.9, p -value <0.05).

Discussion

Intracranial meningiomas are the most common brain tumors. Since first described by Cushing in the early 20th century, there have been significant changes in the management paradigms of this condition. Analyses of the NCDB data showed that observation was the most commonly used modality, followed by surgery, radiotherapy, and combination therapy. Examination of the trends for the period 2004 to 2015 showed an overall increase in the use of observation, with concomitant decline in the use of surgery and/or radiotherapy. A shift toward the use of observation was more pronounced between 2004 and 2009, which continued to be present between 2009 and 2015. Additionally, our analyses showed that surgery was the dominant treatment modality in 2004 and 2005. Agarwal et al demonstrated similar trends in their analysis of the Surveillance, Epidemiology, and End Results (SEER) database for intracranial meningioma between 2004 and 2012. They also described similar results in regard to the breakdown of time period, they noticed that in 2004 and 2005, the use of surgery alone surpassed observation as the dominant treatment modality, and from 2008 to 2012, there was a shift in treatment strategy toward increased use of observation. Interestingly, in 2007, they noticed that the majority of patients underwent either surgery or radiation as a treatment modality, which is different from our findings where observation was the main treatment modality. ¹³ The total number of meningiomas that were diagnosed and treated has increased over time. The number nearly doubled between 2004 and 2015. This increase could be a result of increased detection of tumors because of the increased access to imaging, or possibly due to an increase in the number of CoC-accredited programs reporting newly diagnosed meningioma to the NCDB. Multiple studies have described an increase in the incidence of intracranial meningiomas especially small sized and asymptomatic tumors. ^{14 15 16 17 18} Ambekar et al reported a 39% increase in the annual meningioma case volume in the United States, from 2,823 in 2001 to 3,923 in 2010. ¹⁹ Similarly, Radhakrishnan et al reported an increase in incidence of asymptomatic intracranial meningioma on neuroimaging from 0.16/100,000/year in 1950 to 1969 to 2.28/100,000/year in 1970 to 1989. They attributed the increase in diagnosis due to increased availability and quality of imaging studies. ¹⁸ Similar results have been reported from population studies in other countries, Kuratsu et al reported an incidence rate of asymptomatic intracranial meningioma on neuroimaging of 1.19/100,000/year in their survey in the area of Kumamoto Prefecture in southern Japan between 1989 and 1996, which is higher than studies that were done before the era of computed tomography scan and magnetic resonance imaging. ¹⁶

In our association analyses, we found that patients were more likely to receive surgery or radiation if they were treated at an academic institution in comparison to treatment at a community hospital. These findings correlated with less use of observation as the main treatment modality in academic institutions when compared with community hospitals. This could be due to the availability of more resources and expertise at academic institutions, or due to the complexity (large size, location, or symptoms) of cases seen at academic hospitals in comparison to community hospitals. Another consideration is that academic hospitals are biased to offer more aggressive treatment than community hospitals secondary to a multitude of variables outside of the scope of this study. ^{19 20 21}

For tumors equal to or less than 3 cm in size, observation was the dominant treatment modality, followed by surgery, radiotherapy, and combination therapy. While for tumors more than 3 cm, surgery was the main modality followed by observation, combination therapy, and radiotherapy. This is likely due to the fact that small tumors are more likely to be diagnosed incidentally and are asymptomatic, while larger tumors (> 3 cm) are more likely to be symptomatic with mass effect. The majority of tumors in the surgery group were WHO grade 1. Dolecek et al in their analysis of SEER database included all meningiomas from 2004 to 2011, they examined the initial treatment modality and found that observation was used in 61.8% of WHO grade 1, 29.3% of WHO grade 2, and 31% of WHO grade 3. ²² Agarwal et al reported an increase in the odds for using observation as the main treatment modality for smaller or WHO grade 1 tumors, with an increase in the odds for using surgery for larger or WHO grades 2 and 3 meningiomas. ¹³ For WHO grade 3 tumors, several reports showed that a combination therapy increases survival. ^{23 24 25} Cao et al noted that for WHO grade 3 tumors, no significant difference in progression-free survival between patients with and without radiotherapy. However, radiotherapy was associated with a longer overall survival with a median duration of 89 versus 42 months without radiotherapy. ²³ For patients with intracranial meningiomas and no comorbidities (CDS 0), the main treatment modality was either observation or surgery, which is different from patients with comorbidities where a higher CDS correlated with increased use of observation.

Our mortality analyses showed that regardless of treatment modality, patients treated at academic institutions had improved survival when compared with patients treated at community hospitals. This is likely a result of increased case volume and surgical experience at academic institutions. ^{19 20 21} This phenomenon has been previously demonstrated by Curry et al in their study of craniotomies for meningiomas in the United States between 1988 and 2000, where they reported lower in-hospital mortality for high meningioma case volume hospitals. They also detailed higher mortality rates for the low-volume hospitals at 3.2%, compared with 1.6% at the high-volume hospitals. In addition, their data showed a trend toward lower mortality rates when surgery was performed by surgeons with higher caseloads. ¹² Similar results were reported by other investigators for intracranial meningiomas and other neurosurgical conditions, such as clipping of intracranial aneurysms and treatment of complex brain tumors. ^{19 20 21} We noted an increased mortality in patients with higher comorbidity scores. Ambekar et al reported similar findings through analyzing national inpatient sample database. ¹⁹ Finally, tumor grade strongly correlated with increased mortality which was similar to reports of analyses of other national databases and institutional series. ^{19 26 27}

Limitations

The limitations of our study and the NCDB should be recognized in the light of the above-mentioned results. As the NCDB data represent more than 70% of newly diagnosed cancer cases nationwide and more than 34 million historical records, the database does not include data pertaining to various techniques of surgery, the extent of resection, and anatomical location of the tumor. In addition, NCDB database has information on initial treatment modality. Therefore, all tumors captured by the NCDB are newly diagnosed. The NCDB does not capture information regarding those that failed the primary treatment modality and then switched to a different treatment method. In regard to the grading, this means that all patients with a known tumor grade underwent surgery. In addition, for patients who were treated with observation or radiotherapy as the primary treatment modality, we are unable to discern the frequency of tumor grades in these groups. It also lacks radiotherapy treatment plans including field design. Finally, the data retrieved from the NCDB may be influenced by local treatment biases. Therefore, it may not be entirely representative of the national meningioma population.

Conclusion

The diagnosis of intracranial meningiomas has increased overtime. While the reason for this increase remains elusive, it may be secondary to an increase in the availability of and quality of diagnostic imaging. Moreover, the use of primary observation as a treatment modality has increased relative to other treatment methods, especially for small tumors. Our analyses showed an association between increased mortality and advanced age, higher tumor grade, larger tumor size, and higher comorbidity score. This study provides a unique overview of the evolution of the management of intracranial meningioma in the United States over time.