Abstract
Background Age, tumor size and location, overall health, and patient preference are primary considerations driving treatment decision-making for intracranial meningiomas. However, even for the same individual patient, treatment recommendations may vary between centers and providers.
Objective To study associations between geography, disease presentation, and management of intracranial meningioma in the United States.
Methods The population-based Surveillance, Epidemiology, and End Results(SEER) data were queried between 2004 and 2014 for cases of intracranial meningioma.
Results A total of 65,808 patients with intracranial meningioma were identified. Univariate analyses demonstrated strong associations between geographic location, age, and size of tumor at presentation. The mean age for all registries was 64.2 years, with a range from 62.0 (Utah registry) to 66.6 (Detroit registry). The greatest proportion of small tumors (<1 cm) were identified in the Utah registry (13.9% of tumors), while the greatest proportion of large tumors (> 4cm) were noted in the Hawaii registry (30.7% of tumors). Multivariable analysis demonstrated that the impact of geography on treatment selection was just as important as other established variables. For example, the distribution in tumor size between New Mexico and Greater California registries is nearly identical; however, the odds ratio for surgery was 1.5 times greater for the New Mexico population.
Conclusion These data suggest that disease presentation and treatment are significantly influenced by regional referral patterns, provider or institutional treatment preferences, and regional availability of subspecialty expertise. Understanding such biases is important for patients, referring physicians, and treatment providers in an effort to provide balanced counseling and treatment.
Keywords: skull base, meningioma, intracranial, regional
Introduction
Meningiomas are the most common primary intracranial tumor in the United States. 1 Age, tumor size and location, symptoms, overall health status, and patient preference are primary considerations driving treatment decision-making. 2 However, even for the same individual patient, treatment recommendations may vary between centers and providers based upon availability of radiation platforms, subspecialty expertise, and other less tangible biases. To the best of our knowledge, geographic trends in intracranial meningioma management in the United States have not been evaluated previously.
We sought to evaluate associations between patient geography, disease presentation, and initial treatment. The Surveillance, Epidemiology and End Results (SEER) cancer registries were utilized in conjunction with United States Census Bureau data to evaluate baseline demographic features, tumor characteristics, and treatment strategy by geographic registry and population density. Longitudinal changes pertaining to disease incidence, presentation, and management were explored.
Materials and Methods
The SEER database includes 18 population-based tumor registries covering ∼28% of the United States, from 1973 to 2014. Cancer registrars in the United States began identifying and abstracting benign central nervous system tumors in 2004, including intracranial meningiomas. SEER reporting sources include records from hospitals, hospital-affiliated or independent radiation and medical oncology centers, hospital-affiliated or independent pathology laboratories, hospital outpatient units, outpatient surgery centers, physician offices, nursing or convalescent homes, and autopsy only or death certificate only reports. Internal safeguards within the SEER database system are in place to ensure data quality and reliability across all registries. Cases of intracranial meningioma were included for analysis if their record was confirmed to contain one of the International Classification of Disease for Oncology (ICD-O) histology codes as well as one of the ICD-O topography codes. Topography codes included C70.0, cerebral meninges; C71.0, cerebrum; C71.1, frontal lobe; C71.2, temporal lobe; C71.3, parietal lobe; C71.4, occipital lobe; C71.5, ventricle, NOS; C71.56, cerebellum, NOS; C71.1, brain stem; C71.8, overlapping lesion of the brain; C71.9, brain, NOS; C72.2, olfactory nerve; C72.3, optic nerve; C72.5, cranial nerve, NOS; C75.1, pituitary gland; C75.2, craniopharyngeal duct; C75.3, pineal gland. All 18 SEER registries were reviewed and only those (16 registries) with complete population inclusivity and greater than 150 intracranial meningioma cases 2004 to 2014 were analyzed. A summary of population data for each registry meeting inclusion criteria is included in Table 1 .
Table 1. Population data for 16 SEER registries.
| Registry | Population | Population density (persons/sq. mi) | Incidence (per 100,000 person years) |
|---|---|---|---|
| California (Greater Area excluding San Francisco, Los Angeles, and San Jose | 20,585,610 | 144.4 | 73 |
| California (San Francisco—Oakland) | 4,335,391 | 1756.8 | 81 |
| California (San Jose—Monterey) | 2,514,350 | 391.1 | 69 |
| California (Los Angeles) | 9,818,605 | 2420.5 | 68 |
| Connecticut | 3,574,097 | 738.7 | 58 |
| Michigan (Detroit Area) | 3,863,924 | 1961.9 | 101 |
| Georgia (Greater Area) | 6,195,197 | 118.4 | 73 |
| Georgia (Atlanta Area) | 3,365,297 | 1971.5 | 74 |
| Hawaii | 1,360,301 | 212.3 | 56 |
| Iowa | 3,046,355 | 54.6 | 95 |
| Kentucky | 4,339,367 | 109.5 | 98 |
| Louisiana | 4,533,372 | 104.3 | 71 |
| New Jersey | 8,791,894 | 1186.7 | 62 |
| New Mexico | 2,059,179 | 17.0 | 45 |
| Utah | 2,763,885 | 33.8 | 88 |
| Washington (Seattle—Puget Sound) | 4,590,294 | 260.1 | 131 |
Abbreviation: SEER, Surveillance, Epidemiology, and End Results.
Tumor size was categorized as: 0 to <1 cm, 1 to <2 cm, 2 to <3 cm, 3 to <4 cm, 4+ cm, and missing. Management course was categorized as surgery, radiotherapy, multimodality therapy, or conservative observation. Radiation alone was assigned to those patients who received radiation therapy without microsurgery. If a subject did not receive radiation therapy or microsurgery, they were assigned to the observation cohort. The current study only examined primary management, defined by the initial treatment received within 12 months following diagnosis. If a patient received initial observation, but underwent surgery or radiation treatment within the first 12 months of diagnosis, then surgery or radiation was designated accordingly.
Population counts and land area of all US counties for years 2004 to 2014 were gathered from the US Census Bureau and used to determine population density. Population density calculations were estimated for the 16 SEER registry sites that met inclusion criteria. Given the near constant population density in SEER registry regions over the timeframe of the study, the 2010 population density estimates were used as reference data.
The primary outcome of interest was variance of initial treatment strategy for 0 to 3 cm intracranial meningiomas between the 16 SEER registries. The relationship between each SEER registry population density with age, sex, and tumor size was analyzed as secondary outcomes. Continuous features were summarized with means, medians, and ranges; categorical features were summarized with frequency counts and percentages. Comparisons of baseline patient demographics, tumor size, and treatment strategy between the 16 registries of interest were evaluated using chi-squared tests. Comparison of treatment strategies for 0 to 3 cm intracranial meningiomas within the 16 SEER registries was also evaluated in a multivariable setting after adjusting for differences in patient age and tumor size using logistic regression models. Greater California served as the reference population as it was the largest registry site with characteristics comparable to the mean values from all other registries. Statistical trends by increasing population density for continuous variables, including size and age, were analyzed by linear regression. Categorical variables were tested for statistical trends using the Cochran-Armitage trend test.
The Mayo Clinic Institutional Review Board has deemed analyses of SEER data exempt from review. All analyses were performed using the SAS software package (version 9.4 for Windows; SAS Institute Inc., Cary, North Carolina, United States 2008) and p values <0.05 were considered statistically significant.
Results
In total, 65,808 patients with intracranial meningioma were identified over the time period of 2004 to 2014. The mean age at time of diagnosis was 64 years and 73% were women. Overall, 55% of patients received conservative management, 5% radiation alone, and 40 % microsurgery alone. Details regarding tumor size were available for 53,420 patients. Of these, 9.8 % of tumors were small (0 to <1 cm), 66 % were medium (1 to <3 cm), and 24 % were large (>3 cm).
When analyzing the 16 SEER registries having a minimum of 150 subjects with intracranial meningioma, notable differences in tumor size and treatment were identified. The range of mean age between registries was 4.6 years (62.0–66.6 years), and the proportion of women between registries ranged from 66 to 76 % ( p < 0.001) ( Table 2 ). Kentucky, Seattle, and Utah had the greatest percentage of small (0 to <1 cm) tumors diagnosed, while Hawaii, Los Angeles, and New Mexico had the greatest percentage of large (>3 cm) tumors. Among these 16 sites, trends were identified between increasing population density and tumor size or treatment modality. For example, as population density decreased, there was an increase in tumor size (i.e., for tumors over 3 cm in diameter: 40.3% were in the largest density areas, while 44.0% were in the smallest density areas [ p < 0.0001]). Also noted were trends toward more patients receiving radiation and surgery as population density decreased. Another very notable aspect is the variance in disease incidence between registries. For example, the incidence of tumors in Seattle (131 per 100,000 person years) was nearly three times as high as that of New Mexico (45 per 100,000 person years).
Table 2. Comparison of baseline demographic features and tumor size between 16 SEER registries for patients with intracranial meningioma.
| Variable | Overall ( n = 65,808) | Atlanta ( n = 2503) | Greater CA (excluding SF, LA and SJ) ( n = 14,960) | Connecticut ( n = 2,069) | Detroit ( n = 3,900) | Greater Georgia ( n = 4,537) | Hawaii ( n = 758) | Iowa ( n = 2,898) | Kentucky ( n = 4,255) | Los Angeles ( n = 6,665) | Louisiana ( n = 3,231) | New Jersey ( n = 5,438) | New Mexico ( n = 917) | San Francisco—Oakland ( n = 3,507) | San Jose—Monterey ( n = 1,736) |
Seattle ( n = 6,002) | Utah ( n = 2,432) | p Value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Age | ||||||||||||||||||
| Mean (SD) | 64.2 (16.1) | 62.3 (16.0) | 64.5 (16.1) | 62.3 (16.1) | 66.6 (16.4) | 63.4 (15.7) | 64.3 (16.2) | 65.2 (16.2) | 64.4 (15.9) | 63.9 (16.1) | 63.5 (15.7) | 64.3 (16.2) | 62.8 (16.2) | 65.4 (15.9) | 63.7 (15.5) | 64.7 (15.7) | 62.0 (16.7) | <0.0001 |
| Median | 65.0 | 63.0 | 66.0 | 63.0 | 68.0 | 65.0 | 65.0 | 66.0 | 66.0 | 65.0 | 65.0 | 65.0 | 64.0 | 66.0 | 64.0 | 65.0 | 63.0 | |
| Age group | ||||||||||||||||||
| 0–49 | 13,797 (21.0%) | 595 (23.8%) | 3,057 (20.4%) | 518 (25.0%) | 734 (18.8%) | 976 (21.5%) | 161 (21.2%) | 563 (19.4%) | 861 (20.2%) | 1,461 (21.9%) | 679 (21.0%) | 1,136 (20.9%) | 203 (22.1%) | 661 (18.8%) | 384 (22.1%) | 1,212 (20.2%) | 596 (24.5%) | <0.0001 |
| 50–59 | 11,215 (17.0%) | 483 (19.3%) | 2,513 (16.8%) | 382 (18.5%) | 613 (15.7%) | 752 (16.6%) | 121 (16.0%) | 493 (17.0%) | 714 (16.8%) | 1,094 (16.4%) | 588 (18.2%) | 995 (18.3%) | 167 (18.2%) | 585 (16.7%) | 294 (16.9%) | 999 (16.6%) | 422 (17.4%) | |
| 60+ | 40,796 (62.0%) | 1,425 (56.9%) | 9,390 (62.8%) | 1,169 (56.5%) | 2,553 (65.5%) | 2,809 (61.9%) | 476 (62.8%) | 1,842 (63.6%) | 2,680 (63.0%) | 4,110 (61.7%) | 1,964 (60.8%) | 3,307 (60.8%) | 547 (59.7%) | 2,261 (64.5%) | 1,058 (60.9%) | 3,791 (63.2%) | 1,414 (58.1%) | |
| Sex | ||||||||||||||||||
| Male | 17,480 (26.6%) | 614 (24.5%) | 4,115 (27.5%) | 546 (26.4%) | 991 (25.4%) | 1,260 (27.8%) | 212 (28.0%) | 763 (26.3%) | 1,109 (26.1%) | 1,688 (25.3%) | 896 (27.7%) | 1,521 (28.0%) | 255 (27.8%) | 934 (26.6%) | 484 (27.9%) | 1,503 (25.0%) | 589 (24.2%) | <0.0001 |
| Female | 48,328 (73.4%) | 1,889 (75.5%) | 10,845 (72.5%) | 1,523 (73.6%) | 2,909 (74.6%) | 3,277 (72.2%) | 546 (72.0%) | 2,135 (73.7%) | 3,146 (73.9%) | 4,977 (74.7%) | 2,335 (72.3%) | 3,917 (72.0%) | 662 (72.2%) | 2,573 (73.4%) | 1,252 (72.1%) | 4,499 (75.0%) | 1,843 (75.8%) | |
| Tumor size | ||||||||||||||||||
| Missing | 12,388 | 472 | 2,747 | 550 | 633 | 761 | 126 | 380 | 369 | 1,380 | 867 | 1,552 | 200 | 785 | 373 | 788 | 405 | <0.0001 |
| 0 to <1cm | 5,239 (9.8%) | 199 (9.8%) | 1,070 (8.8%) | 160 (10.5%) | 345 (10.6%) | 365 (9.7%) | 73 (11.6%) | 256 (10.2%) | 524 (13.5%) | 368 (7.0%) | 176 (7.4%) | 323 (8.3%) | 44 (6.1%) | 222 (8.2%) | 113 (8.3%) | 720 (13.8%) | 281 (13.9%) | |
| 1 to <2cm | 16,256 (30.4%) | 550 (27.1%) | 3,571 (29.2%) | 433 (28.5%) | 1,026 (31.4%) | 1,132 (30.0%) | 156 (24.7%) | 837 (33.2%) | 1,382 (35.6%) | 1,370 (25.9%) | 688 (29.1%) | 1,008 (25.9%) | 194 (27.1%) | 770 (28.3%) | 406 (29.8%) | 1,991 (38.2%) | 742 (36.6%) | |
| 2 to <3 cm | 11,474 (21.5%) | 460 (22.6%) | 2,646 (21.7%) | 330 (21.7%) | 768 (23.5%) | 843 (22.3%) | 125 (19.8%) | 587 (23.3%) | 788 (20.3%) | 1,106 (20.9%) | 531 (22.5%) | 837 (21.5%) | 162 (22.6%) | 588 (21.6%) | 283 (20.8%) | 1,033 (19.8%) | 387 (19.1%) | |
| 3 to <4cm | 7,584 (14.2%) | 314 (15.5%) | 1,776 (14.5%) | 221 (14.5%) | 449 (13.7%) | 534 (14.1%) | 84 (13.3%) | 326 (12.9%) | 468 (12.0%) | 874 (16.5%) | 364 (15.4%) | 638 (16.4%) | 116 (16.2%) | 400 (14.7%) | 220 (16.1%) | 574 (11.0%) | 226 (11.1%) | |
| ≥ 4cm | 12,867 (24.1%) | 508 (25.0%) | 3,150 (25.8%) | 375 (24.7%) | 679 (20.8%) | 902 (23.9%) | 194 (30.7%) | 512 (20.3%) | 724 (18.6%) | 1,567 (29.6%) | 605 (25.6%) | 1,080 (27.8%) | 201 (28.0%) | 742 (27.3%) | 341 (25.0%) | 896 (17.2%) | 391 (19.3%) | |
Abbreviation: SD, standard deviation; SEER, Surveillance, Epidemiology, and End Results.
When analyzing the subgroup of patients with tumors between 0 and 3 cm, there were significant differences in primary treatment patterns between regions ( p < 0.0005) ( Table 3 ). Multivariate models were constructed to identify trends regarding surgical decision-making after controlling for tumor size and patient age. With Greater California as the reference, residents of the New Mexico, Los Angeles, and San Jose-Monterey registries had the highest odds of undergoing surgery relative to any treatment. The registries exhibiting the greatest odds of receiving radiation therapy relative to any treatment included New Mexico, Los Angeles, and Greater Georgia. Finally, the registries exhibiting the greatest odds of observation relative to any treatment included Seattle, Metropolitan Detroit, and Iowa ( Table 4 , Fig. 1 ).
Table 3. Overall comparison of initial management strategy for <3 cm intracranial meningioma between 16 SEER registries.
| Variable | Overall ( n = 32,969) | Atlanta ( n = 1,209) | Greater CA (excluding SF, LA, and SJ) ( n = 7,287) | Connecticut ( n = 923) | Detroit ( n = 2,139) | Greater Georgia ( n = 2,340) | Hawaii ( n = 354) | Iowa ( n = 1,680) | Kentucky ( n = 2,694) | Los Angeles ( n = 2,844) | Louisiana ( n = 1,395) | New Jersey ( n = 2,168) | New Mexico ( n = 400) | San Francisco—Oakland ( n = 1,580) | San Jose—Monterey ( n = 802) | Seattle ( n = 3,744) | Utah ( n = 1,410) | p Value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Treatment | ||||||||||||||||||
| Observation | 24,929 (75.6%) | 940 (77.8%) | 5,460 (74.9%) | 629 (68.1%) | 1,730 (80.9%) | 1,723 (73.6%) | 257 (72.6%) | 1,357 (80.8%) | 2,103 (78.1%) | 1,880 (66.1%) | 1,050 (75.3%) | 1,590 (73.3%) | 258 (64.5%) | 1,195 (75.6%) | 562 (70.1%) | 3,129 (83.6%) | 1,066 (75.6%) | <0.0001 |
| Radiation only | 2,052 (6.2%) | 87 (7.2%) | 523 (7.2%) | 76 (8.2%) | 59 (2.8%) | 213 (9.1%) | 44 (12.4%) | 61 (3.6%) | 103 (3.8%) | 271 (9.5%) | 60 (4.3%) | 143 (6.6%) | 50 (12.5%) | 112 (7.1%) | 61 (7.6%) | 157 (4.2%) | 32 (2.3%) | <0.0001 |
| Surgery | 5,988 (18.2%) | 182 (15.1%) | 1,304 (17.9%) | 218 (23.6%) | 350 (16.4%) | 404 (17.3%) | 53 (15.0%) | 262 (15.6%) | 488 (18.1%) | 693 (24.4%) | 285 (20.4%) | 435 (20.1%) | 92 (23.0%) | 273 (17.3%) | 179 (22.3%) | 458 (12.2%) | 312 (22.1%) | <0.0001 |
Abbreviation: SEER, Surveillance, Epidemiology, and End Results.
Table 4. Trends in intracranial meningioma management over time comparing the first and last half of the study period.
| Observation only | Radiation only | Surgery | ||||
|---|---|---|---|---|---|---|
| 2004–2008 | 2009–2014 | 2004–2008 | 2009–2014 | 2004–2008 | 2009–2014 | |
| Overall | 13,185 (49.6%) | 22,934 (58.5%) | 1,565 (5.9%) | 1,913 (4.9%) | 11,823 (44.5%) | 14,388 (36.7%) |
| San Francisco-Oakland SMSA | 701 (49.0%) | 1,155 (55.6%) | 82 (5.7%) | 110 (5.3%) | 647 (45.2%) | 812 (39.1%) |
| Connecticut | 342 (37.8%) | 625 (53.7%) | 75 (8.3%) | 63 (5.4%) | 488 (53.9%) | 476 (40.9%) |
| Metropolitan Detroit | 1,032 (61.5%) | 1,441 (64.9%) | 51 (3.0%) | 66 (3.0%) | 596 (35.5%) | 714 (32.2%) |
| Hawaii | 153 (44.2%) | 229 (55.6%) | 36 (10.4%) | 21 (5.1%) | 157 (45.4%) | 162 (39.3%) |
| Iowa | 700 (61.3%) | 1,110 (63.2%) | 50 (4.4%) | 63 (3.6%) | 392 (34.3%) | 583 (33.2%) |
| New Mexico | 112 (31.4%) | 278 (49.6%) | 20 (5.6%) | 54 (9.6%) | 225 (63.0%) | 228 (40.7%) |
| Seattle (Puget Sound) | 1,447 (63.4%) | 2,615 (70.3%) | 117 (5.1%) | 108 (2.9%) | 719 (31.5%) | 996 (26.8%9 |
| Utah | 429 (49.5%) | 951 (60.8%) | 26 (3.0%) | 38 (2.4%) | 412 (47.5%) | 576 (36.8%) |
| Metropolitan Atlanta | 457 (48.8%) | 996 (63.6%) | 73 (7.8%) | 89 (5.7%) | 406 (43.4%) | 482 (30.8%) |
| San Jose-Monterey | 228 (35.0%) | 590 (54.4%) | 37 (5.7%) | 50 (4.6%) | 386 (59.3%) | 445 (41.0%) |
| Los Angeles | 1,138 (39.8%) | 1,806 (47.4%) | 238 (8.3%0 | 229 (6.0%) | 1,482 (51.9%) | 1,772 (46.6%) |
| Greater California | 2,993 (49.7%) | 5,106 (57.1%) | 374 (6.2%) | 500 (5.6%) | 2,657 (44.1%) | 3,330 (37.3%) |
| Kentucky | 942 (55.1%) | 1,642 (64.5%) | 78 (4.6%) | 84 (3.3%) | 689 (40.3%) | 820 (32.2%) |
| Louisiana | 671 (50.1%) | 989 (52.3%) | 69 (5.2%) | 66 (3.5%) | 600 (44.8%) | 836 (44.2%) |
| New Jersey | 1,061 (45.2%) | 1,692 (54.7%) | 106 (4.5%) | 178 (5.8%) | 1,179 (50.3%) | 1,222 (39.5%) |
| Greater Georgia | 779 (45.8%) | 1,709 (60.2%) | 133 (7.8%) | 194 (6.8%) | 788 (46.4%) | 934 (32.9%) |
Fig. 1.
Proportion of treatment types for intracranial meningioma separated by Surveillance, Epidemiology, and End Results registry from 2004 to 2014.
Additional analyses were performed to evaluate overall and registry-specific management trends over time. When comparing the first half of the study period (2004–2008) to the second half (2009–2014), there was an overall increase in primary observation (49.6–58.5%, p < 0.0001), decrease in surgery (44.5–36.7%, p < 0.0001), and decrease in radiation use (5.9–4.9%, p < 0.0001). However, there was significant variation between registries. For example, there was relatively no change in use of surgery over time in Louisiana in comparison to an overall decrease in surgery across registries. Similarly, an overall decrease in radiation use was seen throughout most registries over this time period, with the exception of New Mexico and New Jersey ( Table 5 ).
Table 5. Trends in intracranial meningioma management over time.
| Observation only | Radiation only | Surgery | ||||
|---|---|---|---|---|---|---|
| 2004–2008 | 2009–2014 | 2004–2008 | 2009–2014 | 2004–2008 | 2009–2014 | |
| Overall | 13,185 (49.6%) | 22,934 (58.5%) | 1,565 (5.9%) | 1,913 (4.9%) | 11,823 (44.5%) | 14,388 (36.7%) |
| San Francisco-Oakland SMSA | 701 (49.0%) | 1,155 (55.6%) | 82 (5.7%) | 110 (5.3%) | 647 (45.2%) | 812 (39.1%) |
| Connecticut | 342 (37.8%) | 625 (53.7%) | 75 (8.3%) | 63 (5.4%) | 488 (53.9%) | 476 (40.9%) |
| Metropolitan Detroit | 1,032 (61.5%) | 1,441 (64.9%) | 51 (3.0%) | 66 (3.0%) | 596 (35.5%) | 714 (32.2%) |
| Hawaii | 153 (44.2%) | 229 (55.6%) | 36 (10.4%) | 21 (5.1%) | 157 (45.4%) | 162 (39.3%) |
| Iowa | 700 (61.3%) | 1,110 (63.2%) | 50 (4.4%) | 63 (3.6%) | 392 (34.3%) | 583 (33.2%) |
| New Mexico | 112 (31.4%) | 278 (49.6%) | 20 (5.6%) | 54 (9.6%) | 225 (63.0%) | 228 (40.7%) |
| Seattle (Puget Sound) | 1,447 (63.4%) | 2,615 (70.3%) | 117 (5.1%) | 108 (2.9%) | 719 (31.5%) | 996 (26.8%9 |
| Utah | 429 (49.5%) | 951 (60.8%) | 26 (3.0%) | 38 (2.4%) | 412 (47.5%) | 576 (36.8%) |
| Metropolitan Atlanta | 457 (48.8%) | 996 (63.6%) | 73 (7.8%) | 89 (5.7%) | 406 (43.4%) | 482 (30.8%) |
| San Jose-Monterey | 228 (35.0%) | 590 (54.4%) | 37 (5.7%) | 50 (4.6%) | 386 (59.3%) | 445 (41.0%) |
| Los Angeles | 1,138 (39.8%) | 1,806 (47.4%) | 238 (8.3%0 | 229 (6.0%) | 1,482 (51.9%) | 1,772 (46.6%) |
| Greater California | 2,993 (49.7%) | 5,106 (57.1%) | 374 (6.2%) | 500 (5.6%) | 2,657 (44.1%0 | 3,330 (37.3%) |
| Kentucky | 942 (55.1%) | 1,642 (64.5%) | 78 (4.6%) | 84 (3.3%) | 689 (40.3%) | 820 (32.2%) |
| Louisiana | 671 (50.1%) | 989 (52.3%) | 69 (5.2%) | 66 (3.5%) | 600 (44.8%) | 836 (44.2%) |
| New Jersey | 1,061 (45.2%) | 1,692 (54.7%) | 106 (4.5%) | 178 (5.8%) | 1,179 (50.3%) | 1,222 (39.5%) |
| Greater Georgia | 779 (45.8%) | 1709 (60.2%) | 133 (7.8%) | 194 (6.8%) | 788 (46.4%) | 934 (32.9%) |
Discussion
Key Results
Patient geography plays a prominent role in disease presentation and treatment selection. The current study noted a significant variability in tumor size at diagnosis, while sex and age distributions were relatively consistent across registries. For example, in Los Angeles, only 7% of intracranial meningiomas were less than 1 cm in size, while 46% were greater than 3 cm. This is compared with Seattle, where 14% of intracranial meningiomas are less than 1 cm in size, while only 28% are greater than 3 cm in size. Theoretical possibilities for such findings include local differences in medical care access and in particular to neuroimaging (e.g., head magnetic resonance imaging or head computed tomography), evaluations leading to coincidental discovery of tumors (migraine headache imaging), racial and socioeconomic disparities, and variance in population education levels. 3 4 5 6 7 8 9 10 11 12 13
The current study also noted a significant difference in treatment modalities utilized between registries for small- to medium-sized intracranial meningiomas, even when accounting for baseline differences in age and tumor size. For example, the distribution in tumor size between New Mexico and Greater California registries is nearly identical; however, the odds ratio for surgery was 1.5 times greater for the New Mexico population. Similarly, the odds ratio for primary observation in (Greater California) was 1.6 times greater than the New Mexico population. These variations likely result from differences in subspecialty access, provider expertise, local referral patterns, and treatment preferences between regional referral centers.
Interpretations
The primary motivations behind treatment selection may vary considerably from patient to patient and are often difficult to fully objectify. Patient factors may include comfort level with surgery or radiation, complication risks, personal experience, and length of convalescence. However, physician counseling is often a significant factor in determining treatment selection rather than underlying patient values or preference.
Previous studies have examined patient decision-making regarding intracranial tumor treatment with multiple available treatment options. A national survey of United Kingdom providers showed variations in primary intracranial meningioma decision-making at differing regional institutions. 14 Pogodzinski et al noted a difference in treatment selection by patients with vestibular schwannoma depending on primary evaluation by a neurotologist, radiosurgeon, or both. 15 Carlson et al showed a strong trend in geographical differences in the management of vestibular schwannomas. 16 A possible explanation for these findings is that provider preference and provider expertise significantly influence treatment recommendations, and ultimately patient choice.
Generalizability
To our knowledge, this is first study to quantify the impact of patient geography on intracranial meningioma management in the United States. Our study shows that place of residence is a significant predictor of treatment selection. Understanding this bias exists is important for patients, referring physicians, and treatment providers.
The strength of the study is highlighted by its inclusion and analysis of nearly 66,000 patients with intracranial meningioma, drawing from a sample population comprising 28% of the US population. The primary limitation of this study is that SEER only catalogues location of residence and does not track location of treatment. While it is assumed that most patients receive care at a large regional referral center, we are unable to track treatment episodes outside an individual's geographic registry. Second, only data concerning disease-specific survival is available; other disease-specific outcomes are not available. Third, while the SEER registry network includes over a quarter of the US population, patient sampling is not random. Instead, SEER has selected 18 registries, which approximate the United States as a whole by varying the geography (i.e., area of the United States and urban/rural representation), although minorities are slightly overrepresented. Relevant to this study, the risk of underreporting is greatest for intracranial meningiomas that are observed by nonhospital outpatient clinics. Despite these limitations, to the authors' knowledge, this is the first study to quantify geographical differences in intracranial meningioma presentation and management in the United States.
Limitations
With regard to the results presented in the current study, SEER registries do not catalogue data concerning regional provider availability and access; however, it is fair to assume that some of the treatment variation seen between registries reflects provider expertise, local referral patterns, and access to subspecialists. For example, some regions may not employ a radiosurgeon and as a result, microsurgery or observation may dominate.
Conclusion
Variations in intracranial meningioma disease presentation and primary management exist throughout the United States. Regional referral patterns, provider or institutional treatment preferences, and regional availability of subspecialty expertise significantly influence treatment decision-making. Understanding such biases is important for patients, referring physicians, and treatment providers in an effort to provide balanced counseling and treatment.
References
- 1.Ostrom Q T, Gittleman H, Fulop Jet al. CBTRUS Statistical Report: primary brain and central nervous system tumors diagnosed in the United States in 2008-2012 Neuro-oncol 201517(Suppl 4):iv1–iv62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Cahill K S, Claus E B. Treatment and survival of patients with nonmalignant intracranial meningioma: results from the Surveillance, Epidemiology, and End Results Program of the National Cancer Institute. Clinical article. J Neurosurg. 2011;115(02):259–267. doi: 10.3171/2011.3.JNS101748. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Amsbaugh M, Ugiliweneza B, Burton E, Skirboll S, Woo S, Boakye M. Patterns of care and outcomes of adjuvant radiotherapy for meningiomas: a surveillance, epidemiology, and end results and Medicare linked analysis. Cureus. 2016;8(04):e567. doi: 10.7759/cureus.567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Dolecek T A, Dressler E V, Thakkar J P, Liu M, Al-Qaisi A, Villano J L. Epidemiology of meningiomas post-Public Law 107-206: The Benign Brain Tumor Cancer Registries Amendment Act. Cancer. 2015;121(14):2400–2410. doi: 10.1002/cncr.29379. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Saraf S, McCarthy B J, Villano J L. Update on meningiomas. Oncologist. 2011;16(11):1604–1613. doi: 10.1634/theoncologist.2011-0193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kaul D, Budach V, Graaf L, Gollrad J, Badakhshi H. Outcome of elderly patients with meningioma after image-guided stereotactic radiotherapy: a study of 100 cases. BioMed Res Int. 2015;2015:868401. doi: 10.1155/2015/868401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chan A K, McGovern R A, Brown L T et al. Disparities in access to deep brain stimulation surgery for Parkinson disease: interaction between African American race and Medicaid use. JAMA Neurol. 2014;71(03):291–299. doi: 10.1001/jamaneurol.2013.5798. [DOI] [PubMed] [Google Scholar]
- 8.Mukherjee D, Zaidi H A, Kosztowski T et al. Disparities in access to neuro-oncologic care in the United States. Arch Surg. 2010;145(03):247–253. doi: 10.1001/archsurg.2009.288. [DOI] [PubMed] [Google Scholar]
- 9.Mukherjee D, Patil C G, Todnem N et al. Racial disparities in Medicaid patients after brain tumor surgery. J Clin Neurosci. 2013;20(01):57–61. doi: 10.1016/j.jocn.2012.05.014. [DOI] [PubMed] [Google Scholar]
- 10.Huo Q, Cai C, Zhang Y et al. Delay in diagnosis and treatment of symptomatic breast cancer in China. Ann Surg Oncol. 2015;22(03):883–888. doi: 10.1245/s10434-014-4076-9. [DOI] [PubMed] [Google Scholar]
- 11.Shapiro M, Chen Q, Huang Qet al. 2016;151(4):338–345 [DOI] [PubMed] [Google Scholar]
- 12.Tatalovich Z, Zhu L, Rolin A, Lewis D R, Harlan L C, Winn D M. Geographic disparities in late stage breast cancer incidence: results from eight states in the United States. Int J Health Geogr. 2015;14:31. doi: 10.1186/s12942-015-0025-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Baade P D, Yu X Q, Smith D P, Dunn J, Chambers S K. Geographic disparities in prostate cancer outcomes--review of international patterns. Asian Pac J Cancer Prev. 2015;16(03):1259–1275. doi: 10.7314/apjcp.2015.16.3.1259. [DOI] [PubMed] [Google Scholar]
- 14.Mohammad M H, Chavredakis E, Zakaria R, Brodbelt A, Jenkinson M D. A national survey of the management of patients with incidental meningioma in the United Kingdom. Br J Neurosurg. 2017;31(04):459–463. doi: 10.1080/02688697.2017.1297765. [DOI] [PubMed] [Google Scholar]
- 15.Pogodzinski M S, Harner S G, Link M J. Patient choice in treatment of vestibular schwannoma. Otolaryngol Head Neck Surg. 2004;130(05):611–616. doi: 10.1016/j.otohns.2004.02.006. [DOI] [PubMed] [Google Scholar]
- 16.Carlson M L, Glasgow A E, Grossardt B R, Habermann E B, Link M J. Does where you live influence how your vestibular schwannoma is managed? Examining geographical differences in vestibular schwannoma treatment across the United States. J Neurooncol. 2016;129(02):269–279. doi: 10.1007/s11060-016-2170-5. [DOI] [PubMed] [Google Scholar]