J Neurooncol DOI 10.1007/s11060-016-2082-4
CLINICAL STUDY
Vestibular schwannoma and pituitary adenoma in the same patient: coincidence or novel clinical association? Matthew L. Carlson1,2 • Neil S. Patel1 • Amy E. Glasgow3 • Elizabeth B. Habermann3 Brandon R. Grossardt4 • Michael J. Link2,1
•
Received: 10 October 2015 / Accepted: 15 February 2016 Ó Springer Science+Business Media New York 2016
Abstract Over the years the authors have evaluated a number of patients with vestibular schwannomas (VS) who have also been diagnosed with a pituitary adenoma (PA). No associations between these tumors have been established to date. The objective of the current study is to investigate the epidemiological association between VS and PA via a population-based study and to supplement these data with a retrospective case series of 12 patients who were evaluated at the authors’ center over the past 15 years. An analysis of the Surveillance, Epidemiology, and End Results database including all patients identified with a diagnosis of VS and/or PA was performed. A comparison between the observed and expected annual incidence was calculated and population differences between those with VS ? PA were compared with single tumor populations. 9888 patients with VS and 26,577 patients with PA were identified among 822.9 million person-years. Within these populations, 31 patients were
diagnosed with both tumor types. Overall, 1 in every 319 patients with VS was also diagnosed with a PA. The average annual incidence for VS was 1.2 per 100,000 persons per year while the average PA rate was 3.2 per 100,000 persons per year. The observed rate of co-incident VS and PA was greater than what is expected by chance alone assuming independence. The cohort of patients with coexisting VS and PA were older and more commonly male compared to VS-only or PA-only groups. These data strongly suggest that a common environmental or genetic predisposition exists for VS and PA development. Further study of this population may help elucidate the cause of tumorigenesis in a subset of patients with seemingly sporadic tumors. Keywords Acoustic neuroma Vestibular schwannoma Pituitary tumor Skull base
Introduction Electronic supplementary material The online version of this article (doi:10.1007/s11060-016-2082-4) contains supplementary material, which is available to authorized users. & Matthew L. Carlson [email protected] 1
Department of Otolaryngology-Head and Neck Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905, USA
2
Department of Neurologic Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905, USA
3
Division of Health Care Policy and Research and the Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, MN 55905, USA
4
Division of Biomedical Statistics and Informatics, Department of Health Sciences Research, College of Medicine, Mayo Clinic, Rochester, MN 55905, USA
Vestibular schwannomas (VS) are rare benign slow growing tumors that arise from the vestibular portion of the eighth cranial nerve. Overall they represent only 8 % of all intracranial tumors and demonstrate an incidence of approximately 1–2 per 100,000 population [1].Pituitary adenomas (PA)represent a heterogeneous group of extraaxial neoplasms that collectively comprise approximately 13 % of all intracranial tumors with an incidence of approximately 3 per 100,000 [2]. Over the years the authors have evaluated a number of patients with sporadic unilateral VS who were also diagnosed with PAs. There are only three possible explanations for concomitant VS and PA development including chance occurrence, environmental influence, or genetic predisposition.
123
J Neurooncol
Given the rare incidence of these tumors, we speculated that the frequency of which we encountered this event was far greater than chance occurrence. Currently there are no known epidemiological or well-characterized genetic associations between VS and PA. Additionally in a large percentage of patients, the precise cause for development of either tumor remains unknown [3–6].The objective of the current study is to characterize an epidemiological association between VS and PA using a large national tumor registry database and to supplement these data with 12 cases evaluated at the authors’ institution over the last 15 years.
Materials and methods Retrospective case series Following Institutional Review Board approval (IRB #15001365), a retrospective chart review was performed at a single tertiary academic referral center. All adult patients who were newly diagnosed with a pituitary tumor and VS, between January 2000 and September 2015, were identified. Data collection included demographic features, age at time of diagnosis, presenting symptomatology, tumor characteristics, and treatment strategy. Hearing capacity and VS tumor size were reported according to the American Academy of Otolaryngology-Head and Neck Surgery consensus guidelines [7].
SEER database analysis The Surveillance, Epidemiology and End Results (SEER) database includes a set of twenty population-based tumor registries covering approximately 28 % of the United States from 1973 to 2012 [8]. VS and PA cases became available in 2004 and were selected using the International Classification of Diseases for Oncology (ICD-O-3) code for benign schwannoma (9560/0) with associated topography code (C72.4) for auditory and vestibular nerve and ICD-O-3 codes for benign pituitary adenoma (8140/0, Adenoma NOS; 8202/0, Microcystic adenoma; 8260/0, Papillary adenoma NOS; 8270/0, Chromophobe adenoma; 8271/0, Prolactinoma; 8272/0, Pituitary adenoma NOS; 8280/0, Acidophil adenoma; 8281/0, Mixed acidophil-basophil adenoma; 8290/0, Oxyphilic adenoma; 8300/0, Basophil adenoma) with associated topographical code (C75.1) for pituitary gland.Patients with alternate coding were excluded. Tumor size was categorized as 0 to\1 cm, 1 to\2 cm, 2 to\3 cm, 3 to\4 cm, C 4 cm, and missing. For the purpose of classifying pituitary tumors, a PA smaller than 1 cm was classified as a microadenoma, and those 1 cm or larger were labeled macroadenoma.
123
As SEER is a set of population-based registries, incidence rates may be calculated. The annual population denominators for the registries (a surrogate for personyears of observation) were taken from SEER-Stat for the calendar years of our specific study, and the observed incidence rates were calculated for VS-only, for PA-only, and for the co-occurrence of VS and PA. Incidence rates are calculated by dividing the number of observed cases of VS and PA by the total number of person-years under observation. Based on the observed incidence rates of VSonly cases and PA-only cases, the expected incidence rate for coexisting VS and PA under the assumption of independence can be calculated as the product of the two separate incidence rates [9]. A 95 % confidence interval (95 % CI) was estimated assuming the observed number of events follows a Poisson distribution and the expected number of events is fixed. Finally, the observed-to-expected ratio (O/E) is calculated by dividing the observed number of cases with coexisting VS and PA by the expected number [9]. This ratio quantifies how frequent these two conditions actually occurred together compared to chance occurrence. For example, an O/E of 10 means these two conditions co-occurred 10 times more frequently than expected assuming independence. Statistical analysis Univariate comparisons of demographic and tumor characteristics were conducted using Chi square tests and student t-tests as appropriate. All analyses were performed using the SASÒ software package (version 9.4; SAS Institute Inc., Cary, NC, USA 2008) and p-values \ 0.05 were considered statistically significant. Institutional Review Board approval was obtained prior to study commencement (15-001365).
Results Retrospective case series Twelve patients met inclusion criteria. Of these, 8 were men and the median age at both VS and PA diagnosis was 59 years (range 36–66yrs) (Fig. 1, Supplemental Fig. 1). In 9 of the 12 patients the diagnosis of VS and PA were made synchronously, while 3 cases were diagnosed sequentially. Of the latter, the diagnosis of VS was made first in 2 cases and PA first in 1 case. The mean time delay between diagnosis of the first and second tumor was 11.6 months (range 0–124 months). Six VS were purely intracanalicular, while 6 had cerebellopontine angle extension. The median cisternal dimension of the latter group was 20.5 mm (range
J Neurooncol
Fig. 1 Axial contrast enhanced T1-weighted MRI demonstrating a left-sided 0.6-cm vestibular schwannoma (white arrow) and a previously treated 1.7-cm pituitary prolactinoma (white arrowhead)
12–23 mm). The median radiologic follow-up was 38 months (range 0–111 months). Four patients underwent stereotactic radiosurgery for treatment of their VS, 2 underwent microsurgical resection, and the remainder were observed with serial MRI. Eight patients had pituitary macroadenomas, defined by a minimum size of 10 mm in greatest dimension. The median PA size was 13 mm (range 4–29 mm). One patient had Cushing’s disease while another had hyperprolactinemia. The rest were diagnosed with nonfunctioning benign adenomas. Four patients were treated with transsphenoidal resection, medical therapy, and/or radiation while the remainder were observed without treatment (Table 1). Over the last 15 years at the authors’ institution an average of approximately 150 patients with a new diagnosis of VS were evaluated annually. Thus, approximately 1 out of every 188 patients with a newly diagnosed VS was also diagnosed with a PA during this time. SEER database analysis Between 2004 and 2012 a total of 9,888 patients with VS, from the SEER registry populations totaling 822.9 million person-years, were reported in the SEER database. The mean age at diagnosis was 54.9 years (SD 14.8) and there were slightly more women than men (52.1 % vs. 47.9 %). The most commonly reported ethnicity was Non-Hispanic White, followed by Hispanic. The observed incidence during this time interval was 1.2 VS per 100,000 personyears (Table 2).
Over this same time period, a total of 26,577 PAs were reported in the SEER database. The mean age at diagnosis was 50.1 years (SD 18.4) and there were 10.6 % more women than men (55.3 % vs. 44.7 %). Similar to the VS population, the most commonly reported ethnicity was Non-Hispanic White, followed by Hispanic. The observed incidence during this time interval was 3.2 PA per 100,000 person-years, approximately 3 times the incidence of VS (Table 3). Between 2004 and 2012, 31 patients (95 % CI 21.1–44.0) with concomitant VS and PA were reported, providing an observed incidence rate of 3.8 9 10-3 (95 % CI 2.6 9 10-3–5.4 9 10-3) cases per 100,000 personyears. The expected rate of having both conditions, determined by calculating the product of the two separate incidence rates, was 3.88 9 10-5 cases per 100,000 person-years. From these data, the calculated O/E ratio was 97.1 (95 % CI 66.0–138) (Table 3). That is, there were 97-fold more cases with both conditions than one would expect under independence of these two conditions. Of the 31 patients with both tumor types, 15 were diagnosed synchronously, whereas 16 cases were diagnosed sequentially. The diagnosis of VS then PA was made in 6 cases and PA followed by VS in 10 cases. The mean time delay between diagnosis of the first and second tumor was 11.3 months (range 0–88 months). On average, every 319 patients with VS also harbored a PA. The average age of this cohort was 59.9 years (SD 13.4);4.8 years greater than the VS-only group (P = 0.07) and 9.8 years greater than the PA only group (P = 0.002). In contrast to the two single-tumor populations, patients with VS and PA were more frequently men (64.5 % vs. 47.9 % in VS group, P = 0.06 and 44.7 % in PA group, P = 0.03). There were no statistically significant differences between groups with regard to tumor size or reported ethnicity. Within the subgroup of 31 patients, 6 had a third tumor type reported, only 1 of which was intracranial (Table 4). There was no difference in the reported distribution of histological subtype between all PAs and the 31 with co-existing VS (Table 5). As an internal control measure, we evaluated the O/E of another relatively common and presumably unrelated intracranial tumor to help as certain whether the increased O/E for VS and PA is valid or simply a reflection of the fact that both conditions are generally diagnosed with head MRI. In other words, because patients with VS and PA both commonly receive multiple MRI studies, it is possible that an incidental tumor of the other type would be found at a greater frequency than in the general population. When evaluating glioblastoma multiforme (GBM; ICD-0-3 9440/3) utilizing identical methodology, an incidence of 2.9 cases per 100,000 person-years was identified. Based on the individual incidences of VS and GBM, the expected
123
J Neurooncol Table 1 Twelve patients with newly diagnosed unilateral vestibular schwannomas and pituitary tumors at the authors’ institution between 2000 and 2015 No.
Demographic data
Vestibular schwannoma data
Pituitary adenoma data
Age at VS Dx
Age at Pit Dx
Sex
Size (mm)a
Side
Hearing classa
Attributable symptoms
Treatment
Size (mm)
Vision changes
Endocrine function
1
41
41
M
IAC
R
A
None
Obs
29
N
Cushing’s
Sur, SRS
2
46
36
M
IAC
L
A
Tin
Obs
17
Y
Prolactinoma
Medical
3
36
37
M
IAC
R
A
HL, Tin
Obs
11
N
–
Obs
4
59
59
F
12
L
A
HL, Imb
SRS
10
N
–
Obs
5
59
59
M
16
L
B
HL, Tin, V, Imb
SRS
15
N
–
Sur
6
66
66
M
20
R
No data
HL, Imb
SRS
15
Y
–
Obs
7
56
56
F
21
R
A
Tin, Imb
Sur
9
N
–
Obs
8
65
65
M
IAC
R
No data
No Data
Obs
29
Y
Apoplexy
Sur
9
63
63
F
23
L
B
HL, Tin, V, Imb
SRS
7
N
–
Obs
10
53
53
F
IAC
L
No data
No Data
No data
4
No data
No data
Obs
11
63
63
M
21
L
D
HL, Tin, Imb
Sur
6
N
–
Obs
12
65
65
M
IAC
L
D
HL, Imb
Obs
20
N
–
Obs
a
Treatment
Size and hearing status reported according to American Academy of Otolaryngology-Head and Neck Surgery guidelines
F, female; HL, hearing loss; IAC, internal auditory canal; Imb, imbalance; L, left; M, male; Obs, observation; Pit, pituitary tumor; R, right; SRS, stereotactic radiosurgery; Sur, surgery; Tin, tinnitus; V, vertigo; VS, vestibular schwannoma
number of cases with both GBM and VS is 0.29, however a total of 5 cases was observed. Thus, the O/E for VS and GBM is 17.5, which is 5.5 times less frequent than the O/E determined for VS and PA. This suggests that coincidental diagnosis of the second tumor by virtue of obtaining a head MRI occurs; however, this alone cannot account for the significantly greater association identified between VS and PA.
Discussion In this report the authors utilize a large national tumor registry to demonstrate a novel epidemiological association between VS and PA and further supplement these findings with a case series of 12 patients that were evaluated at the authors’ center within the last 15 years. To the best of our knowledge, only 4 previous case reports have documented co-occurring VS and PA (Table 6) [10– 13].Currently there are no described syndromes or known familial tendencies that include both VS and PAs within the phenotypic spectrum. Additionally, in a large percentage of patients, the exact cause for development of either tumor remains unknown [3–6].The data in the current study suggest that underlying genetic or environmental factors influence the concomitant development of VS and PA in a subpopulation of patients. If the co-occurrence of these tumors resulted from heritable genetic aberrations, we would expect to see additional cases of affected kin.
123
Among patients with VS, an estimated 5 % are diagnosed with neurofibromatosis type 2 (NF2), an autosomal dominant disease associated with the development of multiple benign intracranial and spinal tumors, excluding pituitary neoplasms. However, data regarding the occurrence of familial non-NF2 VS is sparse. Bikhazi et al. performed a multisite retrospective chart review and accompanying survey questionnaire of Acoustic Neuroma Association members and identified 9 families with 18 individuals having familial unilateral VS. They determined that the occurrence of unilateral VS within families was greater than chance [3].Similarly, Neary and colleagues studied 93 families with at least one member affected by unilateral VS. Of these, 2 families had 2 members affected by a unilateral VS; the first family with affected siblings, and the second with first cousins [4]. Nearly paralleling inheritance patterns of VSs, most PAs are sporadic with 5 % representing pituitary adenomas associated with endocrine-related tumor syndromes such as multiple endocrine neoplasia (types 1 and 4) and Carney’s complex (myxomas and pituitary adenomas causing acromegaly) [14]. Additionally, a distinct group of families containing up to 4 members affected by pituitary adenomas without other endocrine neoplasms emerged in the 1990s. A loss of heterozygosity at the AIP locus (a gene coding for aryl hydrocarbon receptor-interacting protein) was found in 15 % of patients suggesting heritability, but given the diversity in mutations responsible for the remainder, a causative association could not be definitively established [15].
J Neurooncol Table 2 Epidemiologic features of 9857 patients with vestibular scwhannoma, 26,546 with pituitary tumors, and 31 with concomitant vestibular schwannoma and pituitary adenomas Feature
VS only (N = 9857)
PA only (N = 26,546)
Both tumors (N = 31) VS size
Both tumors (N = 31) PA size
Tumor size 2303 (23.4 %)
6512 (24.5 %)
4 (12.9 %)
7 (22.6 %)
0 to \ 1 cm
Missing
1773 (18.0 %)
5498 (20.7 %)
8 (25.8 %)
6 (19.4 %)
1 to \ 2 cm
2844 (28.9 %)
5503 (20.7 %)
11 (35.5 %)
8 (25.8 %)
2 to \ 3 cm
1544 (15.7 %)
5088 (19.2 %)
5 (16.1 %)
6 (19.4 %)
3 to \ 4 cm
810 (8.2 %)
2410 (9.1 %)
2 (6.5 %)
4 (12.9 %)
C 4 cm
583 (5.9 %)
1535 (5.8 %)
1 (3.2 %)
0 (0.0 %)
1060 (10.8 %)
2322 (8.7 %)
0 (0.0 %)
1 (3.2 %)
2005
991 (10.1 %)
2512 (9.5 %)
1 (3.2 %)
0 (0.0 %)
2006
1116 (11.3 %)
2671 (10.1 %)
3 (9.7 %)
3 (9.7 %)
2007
1111 (11.3 %)
2818 (10.6 %)
2 (6.5 %)
1 (3.2 %)
2008
1119 (11.4 %)
2920 (11.0 %)
7 (22.6 %)
6 (19.4 %)
2009 2010
1182 (12.0 %) 1083 (11.0 %)
3209 (12.1 %) 3298 (12.4 %)
4 (12.9 %) 4 (12.9 %)
6 (19.4 %) 5 (16.1 %)
2011
1035 (10.5 %)
3434 (12.9 %)
6 (19.4 %)
4 (12.9 %)
2012
1160 (11.8 %)
3362 (12.7 %)
4 (12.9 %)
5 (16.1 %)
54.9 (14.8)
50.1 (18.4)
Year 2004
Age Mean (SD)
59.9 (13.4)
Under 50
3552 (36.0 %)
13,407 (50.5 %)
8 (25.8 %)
8 (25.8 %)
Age 50–59
2497 (25.3 %)
4461 (16.8 %)
7 (22.6 %)
7 (22.6 %)
16 (51.6 %)
8678 (32.7 %)
3808 (38.6 %)
16 (51.6 %)
Male
4723 (47.9 %)
11,866 (44.7 %)
20 (64.5 %)
20 (64.5 %)
Female
5134 (52.1 %)
14,680 (55.3 %)
11 (35.5 %)
11 (35.5 %)
Gender
Ethnicity 14,525 (54.75)
0.6281
0.7381
0.1820
0.2971
0.0642
0.0018
0.3109
0.0208
0.0727
0.0301
0.2304
0.2872
60.0 (13.6)
Age group
Over 60
VS ? PA versus VS VS ? PA versus PA only (P Value)* only (P-Value)*
Non-hispanic white
7345 (74.5 %)
19 (61.3 %)
19 (61.3 %)
Hispanic Non-hispanic black
821 (8.3 %) 403 (4.1 %)
4395 (16.6 %) 4701 (17.7 %)
6 (19.4 %) 1 (3.2 %)
6 (19.4 %) 1 (3.2 %)
Non-hispanic Asian or Pacific Islander
855 (8.7 %)
1984 (7.5 %)
3 (9.7 %)
3 (9.7 %)
Other or unknown
433 (4.4 %)
941 (3.5 %)
2 (6.5 %)
2 (6.5 %)
PA, pituitary adenoma; VS, vestibular schwannoma * P-values reported are from the Chi square test (categorical variables) and student t test (for the continuous variable age)
The high prevalence and unexpectedly strong familial association of each tumor, combined with the findings of this study, lead to the suspicion that the pathogenesis of VSs and PAs may share a common pathway. One hypothesis is that a common genetic basis underlies the development of both tumors. While at least 20 single-gene mutations have been discovered in sporadic PAs, few other than the NF2 gene have been identified in sporadic VSs [14, 16]. One case report of VS and PAin the same patient
drew attention to fibroblast growth factor receptor 1 (FGFR1) as a potential molecular basis for tumor development. An immunohistochemical study of FGFR1 in PAs demonstrated an inverse correlation between FGFR1 expression and tumor size [17]. On the contrary, however, analysis of 30 sporadic VSs demonstrated a positive correlation between FGFR1 overexpression and growth of VSs [18]. Since current descriptions of coexistent VS and PAare limited to case reports, a meaningful genetic
123
J Neurooncol Table 3 Observed and expected incidence calculations for vestibular schwannoma and pituitary adenomas Observed VS incidencea
Observed PA incidencea
Observed VS ? PA incidencea
Expected VS ? PA incidencea
Observed: expected rate ratio
88,254,554
1.201,071,165
2.6321589
0
0.00003161
0
88,929,243
1.115493584
2.8471762
0.00112449
0.00003151
35.68727
3 2
89,688,379 90,542,359
1.247653277 1.229258893
2.98143419 3.11345986
0.003344915 0.002208911
0.00003720 0.00003827
89.922 57.71538
2926
7
91,469,224
1.231015145
3.19889015
0.007652847
0.00003938
194.3392
3215
4
92,325,435
1.28458642
3.48224734
0.004332501
0.00004473
96.85358
1087
3303
4
93,158,638
1.166826849
3.54556493
0.004293751
0.00004137
103.7875
2011
1041
3438
6
93,886,873
1.108781203
3.66185377
0.00639067
0.00004060
157.3981
2012
1164
3367
4
94,620,901
1.230172179
3.55841042
0.004227396
0.00004377
96.57194
All years
9888
26,577
31
822,875,606
1.201639704
3.2297712
0.003767277
0.00003881
97.06920385
Year
# VS
# PA
# PA ? V S
2004
1060
2323
0
2005
992
2512
1
2006 2007
1119 1113
2674 2819
2008
1126
2009
1186
2010
Total SEER population
PA, pituitary adenoma; VS, vestibular schwannoma a
Per 100,000 persons
Table 4 Additional tumors identified in subgroup of 31 patients with coexisting vestibular schwannomas and pituitary adenomas Cancer
N (%)
Body of stomach
1 (3.2 %)
Upper lobe, lung
1 (3.2 %)
Bone marrow
1 (3.2 %)
Upper-outer quadrant of breast (UOQ)
1 (3.2 %)
Prostate gland; Prostate NOS
1 (3.2 %)
Cerebral Meninges
1 (3.2 %)
analysis of tumors in patients who manifest both has yet to be performed. The coprevalence data presented in this study may justify future tumor genome sequencing in a series of patients with coexistent VS and PA. An alternative hypothesis for the strong familiality and coprevalence of VSs and PAs is that a common environmental trigger or exposure may underlie the development of both. No known risk factors exist for the development of
Table 5 Histological subtypes of pituitary adenomas reported overall and within the select group of patients having coexisting vestibular schwannoma and pituitary adenoma
PAs in the absence of a known syndrome or mutation. Some studies have demonstrated associations between radiation exposure, noise exposure, and personal history of chickenpox and the development of VS [19]. The link between local radiofrequency exposure from cell phones and the development of VS was demonstrated in a subgroup of patients with greater than 10 years of cumulative cell phone use (particularly in the ipsilateral ear) in the Swedish INTERPHONE study [20]. A larger follow up study with more long-term cell phone users failed to show a positive correlation, however, this study instead demonstrated that patients frequently switched their preferred side of cell phone use long before the diagnosis of VS was made [21]. These data support the notion that cell phone use may increase the likelihood of detecting hearing loss as the presenting symptom of VS, but fail to convincingly defend the hypothesis that cell phones lead to their development. Postmortem histological studies regularly demonstrate that the true prevalence of many benign and malignant
ICD-0-3 Code: pituitary adenoma subtype 8140/0: Adenoma, NOS
Patients with VS ? PA
2193 (8.3 %)
1 (3.2 %)
8202/0: Microcystic adenoma
3 (0.0 %)
0 (0.0 %)
8260/0: Papillary adenoma, NOS
9 (0.0 %)
0 (0.0 %)
242 (0.9 %) 957 (3.6 %)
1 (3.2 %) 2 (6.5 %)
8270/0: Chromophobe adenoma 8271/0: Prolactinoma 8272/0: Pituitary adenoma, NOS 8280/0: Acidophil adenoma 8281/0: Mixed acidophil-basophil adenoma
123
All patients with PA
23,117 (87.0 %)
27 (87.1 %)
33 (0.1 %)
0 (0.0 %)
1 (0.0 %)
0 (0.0 %)
8290/0: Oxyphilic adenoma
16 (0.1 %)
0 (0.0 %)
8300/0: Basophil adenoma
6 (0.0 %)
0 (0.0 %)
Surgery Not specified
b
a
L Largeb F 28 Gorman and Hewer (1984) 4
Size was not specified, although ‘‘evidence of brain stem compression and obliteration of the IVth ventricle’’ was reported Size was not specified, although ‘‘large left acoustic neuroma complicated by hydrocephalus’’ was reported
Not specified Not specified Surgery HL, Imb
N ‘‘Macroadenoma’’ SRS None None L Vaamonde Lago et al. [13] 3
23
F
14
‘‘Nerve deafness’’
SRS and medical Medical Elevated growth hormone Elevated Prolactin N 16 SRS Imb, Tin ‘‘Severe’’ L 22 F Niu et al. [11] 2
42
Surgery Mild elevated Prolactin Y 40 Obs Imb L Amit et al. [12] 1
70
M
Largea
‘‘Slight’’
Vision changes Size (mm) Treatment Attributable symptoms Hearing loss Side Size (mm)a Publication
Age
Sex
Vestibular schwannoma data Demographic data No.
Table 6 Summary of 4 case reports describing the co-occurrence of vestibular schwannoma and pituitary adenoma
Pituitary adenoma data
Endocrine function
Treatment
J Neurooncol
tumors is much greater than what has been reported by clinical studies, suggesting that a portion of the population harbor subclinical disease [14, 22]. We have also observed that the clinical prevalence of many benign tumors, including VSs and PAs, has also increased secondary to more sensitive radiologic studies and greater patient access to advanced imaging [14, 23]. These phenomena, however, cannot account for the greater co-incidence of VS and PA observed in the current study for several reasons. Theoretically, any inherent biases or limitations of the SEER database should be cancelled since data from both patient populations were obtained from the same registry and our analysis utilized the incidence ratio of observed-to-expected. Additionally, patients catalogued in the SEER database presumably received appropriate imaging to establish an accurate diagnosis, and based on the study period from 2004 to 2012, the great majority of patients likely underwent MRI with contrast—the imaging modality of choice for diagnosis of both VS and PAs. Furthermore, only approximately one-fifth of all VS and PA reported in this series were less than 1 cm, and thus it is fair to assume that a significant portion were symptomatic and not simply small incidental tumors. The tumor characteristics for the 31 patients with both VS and PA were not statistically different from the population of patients with single primary tumors (Table 2). Lastly, as an internal control measure, we evaluated the co-occurrence of VS and another intracranial tumor to ensure our findings were not simply the result of methodological error or coincidental secondary diagnosis from head MRI. We found that the co-occurrence of GBM and VS in the SEER database was 17.5. Assuming that the development of VS and GBM are independent events, this suggests that a certain number of incidental tumors are diagnosed by virtue of obtaining a head MRI. However, the O/E identified for VS and GBM was still 5.5 times less than the O/E of VS and PA, further supporting a true association. In closing, there are several limitations inherent to study design that warrant consideration. The SEER database reports aggregate data from 20 regional US cancer registries, representing a selected sample of approximately 28 % of the US population. Therefore, the SEER population may not be entirely representative of the national population. Second, patient-and tumor-related variables collected in the SEER database are generic and are not tailored to the study of VS or PA. Thus many important disease-specific features within these populations are not reported. Last, beyond internal quality control measures, there is no method of SEER data verification—our analysis was limited by the quality and completeness of reporting by the individual registries. Despite these shortcomings, the SEER database affords an invaluable tool for analysis of national epidemiological trends that is particularly
123
J Neurooncol
useful for the study of rare conditions such as coexistent VS and PA.
Conclusions Using a large population based tumor registry, the current study demonstrates that the observed rate of co-incident VS and PA is greater than what is expected by chance alone. These data suggest that a common environmental or genetic predisposition exists for VS and PA development. Further study of this population may help elucidate the cause of tumorigenesis in a subset of patients with seemingly sporadic tumors.
Financial material and support Internal departmental funding was utilized without commercial sponsorship or support. Compliance with ethical standards Conflict of interest
None.
References 1. Carlson ML, Habermann EB, Wagie AE, Driscoll CL, Van Gompel JJ, Jacob JT, Link MJ (2015) The Changing Landscape of Vestibular Schwannoma Management in the United States-A Shift Toward Conservatism. Otolaryngol Head Neck Surg. doi:10.1177/0194599815590105 2. McDowell BD, Wallace RB, Carnahan RM, Chrischilles EA, Lynch CF, Schlechte JA (2011) Demographic differences in incidence for pituitary adenoma. Pituitary 14:23–30. doi:10.1007/ s11102-010-0253-4 3. Bikhazi NB, Slattery WH 3rd, Lalwani AK, Jackler RK, Bikhazi PH, Brackmann DE (1997) Familial occurrence of unilateral vestibular schwannoma. Laryngoscope 107:1176–1180 4. Neary WJ, Newton VE, Laoide-Kemp SN, Ramsden RT, Griffith G, Evans DG, Harris R, Strachan T (1996) A clinical, genetic and audiological study of patients and families with unilateral vestibular schwannomas. I. Clinical features of neurofibromatosis in patients with unilateral vestibular schwannomas. J Larnyngol Otol 110:634–640 5. Dworakowska D, Grossman AB (2009) The pathophysiology of pituitary adenomas. Best Pract Res Clin Endocrinol Metab 23:525–541. doi:10.1016/j.beem.2009.05.004 6. Couldwell WT, Cannon-Albright L (2010) A heritable predisposition to pituitary tumors. Pituitary 13:130–137. doi:10.1007/ s11102-009-0212-0 7. Committee on Hearing and Equilibrium guidelines for the evaluation of hearing preservation in acoustic neuroma (vestibular schwannoma) (1995) Otolaryngol Head Neck Surg 113:179–180 8. Overview of the SEER Program. (n.d.). Retrieved March 29, from http://seer.cancer.gov/about/overview.html
123
9. Selvin S (2004) Statistical analysis of epidemiology data. Oxford University Press, Oxford 10. Couldwell WT, Cannon-Albright LA (2014) Co-prevalence of other tumors in patients harboring pituitary tumors. J Neurosurg 121:1474–1477. doi:10.3171/2014.8.JNS132139 11. Niu Y, Ma L, Mao Q, Wu L, Chen J (2010) Pituitary adenoma and vestibular schwannoma: case report and review of the literature. J Postgrad Med 56:281–283. doi:10.4103/0022-3859.70940 12. Amit A, Achawal S, Dorward N (2008) Pituitary macro adenoma and vestibular schwannoma: a case report of dual intracranial pathologies. Br J Neurosurg 22:695–696. doi:10.1080/02688690 801987366 13. Vaamonde Lago P, Castro Vilas C, Soto Varela A, Frade Gonzalez C, Santos Perez S, Labella Caballero T (2001) Asymptomatic acoustic neurinoma associated with hypophyseal macroadenoma. Acta Otorrinolaringol Esp 52:705–708 14. Daly AF, Tichomirowa MA, Beckers A (2009) The epidemiology and genetics of pituitary adenomas. Best Pract Res Clin Endocrinol Metab 23:543–554. doi:10.1016/j.beem.2009.05.008 15. Daly AF, Vanbellinghen JF, Khoo SK, Jaffrain-Rea ML, Naves LA, Guitelman MA, Murat A, Emy P, Gimenez-Roqueplo AP, Tamburrano G, Raverot G, Barlier A, De Herder W, Penfornis A, Ciccarelli E, Estour B, Lecomte P, Gatta B, Chabre O, Sabate MI, Bertagna X, Garcia Basavilbaso N, Stalldecker G, Colao A, Ferolla P, Wemeau JL, Caron P, Sadoul JL, Oneto A, Archambeaud F, Calender A, Sinilnikova O, Montanana CF, Cavagnini F, Hana V, Solano A, Delettieres D, Luccio-Camelo DC, Basso A, Rohmer V, Brue T, Bours V, Teh BT, Beckers A (2007) Aryl hydrocarbon receptor-interacting protein gene mutations in familial isolated pituitary adenomas: analysis in 73 families. J Clin Endocrinol Metab 92:1891–1896. doi:10.1210/jc.20062513 16. Mohyuddin A, Neary WJ, Wallace A, Wu CL, Purcell S, Reid H, Ramsden RT, Read A, Black G, Evans DG (2002) Molecular genetic analysis of the NF2 gene in young patients with unilateral vestibular schwannomas. J Med Genet 39:315–322 17. Fukui S, Otani N, Nawashiro H, Yano A, Nomura N, Miyazawa T, Ohnuki A, Tsuzuki N, Katoh H, Ishihara S, Shima K (2002) Subcellular localization of basic fibroblast growth factor and fibroblast growth factor receptor 1 in pituitary adenomas. Brain tumor Pathol 19:23–29 18. O’Reilly BF, Kishore A, Crowther JA, Smith C (2004) Correlation of growth factor receptor expression with clinical growth in vestibular schwannomas. Otol Neurotol 25:791–796 19. Corona AP, Ferrite S, Lopes Mda S, Rego MA (2012) Risk factors associated with vestibular nerve schwannomas. Otol Neurotol 33:459–465. doi:10.1097/MAO.0b013e3182487fee 20. Lonn S, Ahlbom A, Hall P, Feychting M (2004) Mobile phone use and the risk of acoustic neuroma. Epidemiology 15:653–659 21. Pettersson D, Mathiesen T, Prochazka M, Bergenheim T, Florentzson R, Harder H, Nyberg G, Siesjo P, Feychting M (2014) Long-term mobile phone use and acoustic neuroma risk. Epidemiology 25:233–241. doi:10.1097/EDE.0000000000000058 22. Stewart TJ, Liland J, Schuknecht HF (1975) Occult schwannomas of the vestibular nerve. Arch Otolaryngol 101:91–95 23. Stangerup SE, Tos M, Thomsen J, Caye-Thomasen P (2010) True incidence of vestibular schwannoma? Neurosurgery 67:1335–1340. doi:10.1227/NEU.0b013e3181f22660 discussion 1340
CLINICAL STUDY
Vestibular schwannoma and pituitary adenoma in the same patient: coincidence or novel clinical association? Matthew L. Carlson1,2 • Neil S. Patel1 • Amy E. Glasgow3 • Elizabeth B. Habermann3 Brandon R. Grossardt4 • Michael J. Link2,1
•
Received: 10 October 2015 / Accepted: 15 February 2016 Ó Springer Science+Business Media New York 2016
Abstract Over the years the authors have evaluated a number of patients with vestibular schwannomas (VS) who have also been diagnosed with a pituitary adenoma (PA). No associations between these tumors have been established to date. The objective of the current study is to investigate the epidemiological association between VS and PA via a population-based study and to supplement these data with a retrospective case series of 12 patients who were evaluated at the authors’ center over the past 15 years. An analysis of the Surveillance, Epidemiology, and End Results database including all patients identified with a diagnosis of VS and/or PA was performed. A comparison between the observed and expected annual incidence was calculated and population differences between those with VS ? PA were compared with single tumor populations. 9888 patients with VS and 26,577 patients with PA were identified among 822.9 million person-years. Within these populations, 31 patients were
diagnosed with both tumor types. Overall, 1 in every 319 patients with VS was also diagnosed with a PA. The average annual incidence for VS was 1.2 per 100,000 persons per year while the average PA rate was 3.2 per 100,000 persons per year. The observed rate of co-incident VS and PA was greater than what is expected by chance alone assuming independence. The cohort of patients with coexisting VS and PA were older and more commonly male compared to VS-only or PA-only groups. These data strongly suggest that a common environmental or genetic predisposition exists for VS and PA development. Further study of this population may help elucidate the cause of tumorigenesis in a subset of patients with seemingly sporadic tumors. Keywords Acoustic neuroma Vestibular schwannoma Pituitary tumor Skull base
Introduction Electronic supplementary material The online version of this article (doi:10.1007/s11060-016-2082-4) contains supplementary material, which is available to authorized users. & Matthew L. Carlson [email protected] 1
Department of Otolaryngology-Head and Neck Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905, USA
2
Department of Neurologic Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905, USA
3
Division of Health Care Policy and Research and the Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, MN 55905, USA
4
Division of Biomedical Statistics and Informatics, Department of Health Sciences Research, College of Medicine, Mayo Clinic, Rochester, MN 55905, USA
Vestibular schwannomas (VS) are rare benign slow growing tumors that arise from the vestibular portion of the eighth cranial nerve. Overall they represent only 8 % of all intracranial tumors and demonstrate an incidence of approximately 1–2 per 100,000 population [1].Pituitary adenomas (PA)represent a heterogeneous group of extraaxial neoplasms that collectively comprise approximately 13 % of all intracranial tumors with an incidence of approximately 3 per 100,000 [2]. Over the years the authors have evaluated a number of patients with sporadic unilateral VS who were also diagnosed with PAs. There are only three possible explanations for concomitant VS and PA development including chance occurrence, environmental influence, or genetic predisposition.
123
J Neurooncol
Given the rare incidence of these tumors, we speculated that the frequency of which we encountered this event was far greater than chance occurrence. Currently there are no known epidemiological or well-characterized genetic associations between VS and PA. Additionally in a large percentage of patients, the precise cause for development of either tumor remains unknown [3–6].The objective of the current study is to characterize an epidemiological association between VS and PA using a large national tumor registry database and to supplement these data with 12 cases evaluated at the authors’ institution over the last 15 years.
Materials and methods Retrospective case series Following Institutional Review Board approval (IRB #15001365), a retrospective chart review was performed at a single tertiary academic referral center. All adult patients who were newly diagnosed with a pituitary tumor and VS, between January 2000 and September 2015, were identified. Data collection included demographic features, age at time of diagnosis, presenting symptomatology, tumor characteristics, and treatment strategy. Hearing capacity and VS tumor size were reported according to the American Academy of Otolaryngology-Head and Neck Surgery consensus guidelines [7].
SEER database analysis The Surveillance, Epidemiology and End Results (SEER) database includes a set of twenty population-based tumor registries covering approximately 28 % of the United States from 1973 to 2012 [8]. VS and PA cases became available in 2004 and were selected using the International Classification of Diseases for Oncology (ICD-O-3) code for benign schwannoma (9560/0) with associated topography code (C72.4) for auditory and vestibular nerve and ICD-O-3 codes for benign pituitary adenoma (8140/0, Adenoma NOS; 8202/0, Microcystic adenoma; 8260/0, Papillary adenoma NOS; 8270/0, Chromophobe adenoma; 8271/0, Prolactinoma; 8272/0, Pituitary adenoma NOS; 8280/0, Acidophil adenoma; 8281/0, Mixed acidophil-basophil adenoma; 8290/0, Oxyphilic adenoma; 8300/0, Basophil adenoma) with associated topographical code (C75.1) for pituitary gland.Patients with alternate coding were excluded. Tumor size was categorized as 0 to\1 cm, 1 to\2 cm, 2 to\3 cm, 3 to\4 cm, C 4 cm, and missing. For the purpose of classifying pituitary tumors, a PA smaller than 1 cm was classified as a microadenoma, and those 1 cm or larger were labeled macroadenoma.
123
As SEER is a set of population-based registries, incidence rates may be calculated. The annual population denominators for the registries (a surrogate for personyears of observation) were taken from SEER-Stat for the calendar years of our specific study, and the observed incidence rates were calculated for VS-only, for PA-only, and for the co-occurrence of VS and PA. Incidence rates are calculated by dividing the number of observed cases of VS and PA by the total number of person-years under observation. Based on the observed incidence rates of VSonly cases and PA-only cases, the expected incidence rate for coexisting VS and PA under the assumption of independence can be calculated as the product of the two separate incidence rates [9]. A 95 % confidence interval (95 % CI) was estimated assuming the observed number of events follows a Poisson distribution and the expected number of events is fixed. Finally, the observed-to-expected ratio (O/E) is calculated by dividing the observed number of cases with coexisting VS and PA by the expected number [9]. This ratio quantifies how frequent these two conditions actually occurred together compared to chance occurrence. For example, an O/E of 10 means these two conditions co-occurred 10 times more frequently than expected assuming independence. Statistical analysis Univariate comparisons of demographic and tumor characteristics were conducted using Chi square tests and student t-tests as appropriate. All analyses were performed using the SASÒ software package (version 9.4; SAS Institute Inc., Cary, NC, USA 2008) and p-values \ 0.05 were considered statistically significant. Institutional Review Board approval was obtained prior to study commencement (15-001365).
Results Retrospective case series Twelve patients met inclusion criteria. Of these, 8 were men and the median age at both VS and PA diagnosis was 59 years (range 36–66yrs) (Fig. 1, Supplemental Fig. 1). In 9 of the 12 patients the diagnosis of VS and PA were made synchronously, while 3 cases were diagnosed sequentially. Of the latter, the diagnosis of VS was made first in 2 cases and PA first in 1 case. The mean time delay between diagnosis of the first and second tumor was 11.6 months (range 0–124 months). Six VS were purely intracanalicular, while 6 had cerebellopontine angle extension. The median cisternal dimension of the latter group was 20.5 mm (range
J Neurooncol
Fig. 1 Axial contrast enhanced T1-weighted MRI demonstrating a left-sided 0.6-cm vestibular schwannoma (white arrow) and a previously treated 1.7-cm pituitary prolactinoma (white arrowhead)
12–23 mm). The median radiologic follow-up was 38 months (range 0–111 months). Four patients underwent stereotactic radiosurgery for treatment of their VS, 2 underwent microsurgical resection, and the remainder were observed with serial MRI. Eight patients had pituitary macroadenomas, defined by a minimum size of 10 mm in greatest dimension. The median PA size was 13 mm (range 4–29 mm). One patient had Cushing’s disease while another had hyperprolactinemia. The rest were diagnosed with nonfunctioning benign adenomas. Four patients were treated with transsphenoidal resection, medical therapy, and/or radiation while the remainder were observed without treatment (Table 1). Over the last 15 years at the authors’ institution an average of approximately 150 patients with a new diagnosis of VS were evaluated annually. Thus, approximately 1 out of every 188 patients with a newly diagnosed VS was also diagnosed with a PA during this time. SEER database analysis Between 2004 and 2012 a total of 9,888 patients with VS, from the SEER registry populations totaling 822.9 million person-years, were reported in the SEER database. The mean age at diagnosis was 54.9 years (SD 14.8) and there were slightly more women than men (52.1 % vs. 47.9 %). The most commonly reported ethnicity was Non-Hispanic White, followed by Hispanic. The observed incidence during this time interval was 1.2 VS per 100,000 personyears (Table 2).
Over this same time period, a total of 26,577 PAs were reported in the SEER database. The mean age at diagnosis was 50.1 years (SD 18.4) and there were 10.6 % more women than men (55.3 % vs. 44.7 %). Similar to the VS population, the most commonly reported ethnicity was Non-Hispanic White, followed by Hispanic. The observed incidence during this time interval was 3.2 PA per 100,000 person-years, approximately 3 times the incidence of VS (Table 3). Between 2004 and 2012, 31 patients (95 % CI 21.1–44.0) with concomitant VS and PA were reported, providing an observed incidence rate of 3.8 9 10-3 (95 % CI 2.6 9 10-3–5.4 9 10-3) cases per 100,000 personyears. The expected rate of having both conditions, determined by calculating the product of the two separate incidence rates, was 3.88 9 10-5 cases per 100,000 person-years. From these data, the calculated O/E ratio was 97.1 (95 % CI 66.0–138) (Table 3). That is, there were 97-fold more cases with both conditions than one would expect under independence of these two conditions. Of the 31 patients with both tumor types, 15 were diagnosed synchronously, whereas 16 cases were diagnosed sequentially. The diagnosis of VS then PA was made in 6 cases and PA followed by VS in 10 cases. The mean time delay between diagnosis of the first and second tumor was 11.3 months (range 0–88 months). On average, every 319 patients with VS also harbored a PA. The average age of this cohort was 59.9 years (SD 13.4);4.8 years greater than the VS-only group (P = 0.07) and 9.8 years greater than the PA only group (P = 0.002). In contrast to the two single-tumor populations, patients with VS and PA were more frequently men (64.5 % vs. 47.9 % in VS group, P = 0.06 and 44.7 % in PA group, P = 0.03). There were no statistically significant differences between groups with regard to tumor size or reported ethnicity. Within the subgroup of 31 patients, 6 had a third tumor type reported, only 1 of which was intracranial (Table 4). There was no difference in the reported distribution of histological subtype between all PAs and the 31 with co-existing VS (Table 5). As an internal control measure, we evaluated the O/E of another relatively common and presumably unrelated intracranial tumor to help as certain whether the increased O/E for VS and PA is valid or simply a reflection of the fact that both conditions are generally diagnosed with head MRI. In other words, because patients with VS and PA both commonly receive multiple MRI studies, it is possible that an incidental tumor of the other type would be found at a greater frequency than in the general population. When evaluating glioblastoma multiforme (GBM; ICD-0-3 9440/3) utilizing identical methodology, an incidence of 2.9 cases per 100,000 person-years was identified. Based on the individual incidences of VS and GBM, the expected
123
J Neurooncol Table 1 Twelve patients with newly diagnosed unilateral vestibular schwannomas and pituitary tumors at the authors’ institution between 2000 and 2015 No.
Demographic data
Vestibular schwannoma data
Pituitary adenoma data
Age at VS Dx
Age at Pit Dx
Sex
Size (mm)a
Side
Hearing classa
Attributable symptoms
Treatment
Size (mm)
Vision changes
Endocrine function
1
41
41
M
IAC
R
A
None
Obs
29
N
Cushing’s
Sur, SRS
2
46
36
M
IAC
L
A
Tin
Obs
17
Y
Prolactinoma
Medical
3
36
37
M
IAC
R
A
HL, Tin
Obs
11
N
–
Obs
4
59
59
F
12
L
A
HL, Imb
SRS
10
N
–
Obs
5
59
59
M
16
L
B
HL, Tin, V, Imb
SRS
15
N
–
Sur
6
66
66
M
20
R
No data
HL, Imb
SRS
15
Y
–
Obs
7
56
56
F
21
R
A
Tin, Imb
Sur
9
N
–
Obs
8
65
65
M
IAC
R
No data
No Data
Obs
29
Y
Apoplexy
Sur
9
63
63
F
23
L
B
HL, Tin, V, Imb
SRS
7
N
–
Obs
10
53
53
F
IAC
L
No data
No Data
No data
4
No data
No data
Obs
11
63
63
M
21
L
D
HL, Tin, Imb
Sur
6
N
–
Obs
12
65
65
M
IAC
L
D
HL, Imb
Obs
20
N
–
Obs
a
Treatment
Size and hearing status reported according to American Academy of Otolaryngology-Head and Neck Surgery guidelines
F, female; HL, hearing loss; IAC, internal auditory canal; Imb, imbalance; L, left; M, male; Obs, observation; Pit, pituitary tumor; R, right; SRS, stereotactic radiosurgery; Sur, surgery; Tin, tinnitus; V, vertigo; VS, vestibular schwannoma
number of cases with both GBM and VS is 0.29, however a total of 5 cases was observed. Thus, the O/E for VS and GBM is 17.5, which is 5.5 times less frequent than the O/E determined for VS and PA. This suggests that coincidental diagnosis of the second tumor by virtue of obtaining a head MRI occurs; however, this alone cannot account for the significantly greater association identified between VS and PA.
Discussion In this report the authors utilize a large national tumor registry to demonstrate a novel epidemiological association between VS and PA and further supplement these findings with a case series of 12 patients that were evaluated at the authors’ center within the last 15 years. To the best of our knowledge, only 4 previous case reports have documented co-occurring VS and PA (Table 6) [10– 13].Currently there are no described syndromes or known familial tendencies that include both VS and PAs within the phenotypic spectrum. Additionally, in a large percentage of patients, the exact cause for development of either tumor remains unknown [3–6].The data in the current study suggest that underlying genetic or environmental factors influence the concomitant development of VS and PA in a subpopulation of patients. If the co-occurrence of these tumors resulted from heritable genetic aberrations, we would expect to see additional cases of affected kin.
123
Among patients with VS, an estimated 5 % are diagnosed with neurofibromatosis type 2 (NF2), an autosomal dominant disease associated with the development of multiple benign intracranial and spinal tumors, excluding pituitary neoplasms. However, data regarding the occurrence of familial non-NF2 VS is sparse. Bikhazi et al. performed a multisite retrospective chart review and accompanying survey questionnaire of Acoustic Neuroma Association members and identified 9 families with 18 individuals having familial unilateral VS. They determined that the occurrence of unilateral VS within families was greater than chance [3].Similarly, Neary and colleagues studied 93 families with at least one member affected by unilateral VS. Of these, 2 families had 2 members affected by a unilateral VS; the first family with affected siblings, and the second with first cousins [4]. Nearly paralleling inheritance patterns of VSs, most PAs are sporadic with 5 % representing pituitary adenomas associated with endocrine-related tumor syndromes such as multiple endocrine neoplasia (types 1 and 4) and Carney’s complex (myxomas and pituitary adenomas causing acromegaly) [14]. Additionally, a distinct group of families containing up to 4 members affected by pituitary adenomas without other endocrine neoplasms emerged in the 1990s. A loss of heterozygosity at the AIP locus (a gene coding for aryl hydrocarbon receptor-interacting protein) was found in 15 % of patients suggesting heritability, but given the diversity in mutations responsible for the remainder, a causative association could not be definitively established [15].
J Neurooncol Table 2 Epidemiologic features of 9857 patients with vestibular scwhannoma, 26,546 with pituitary tumors, and 31 with concomitant vestibular schwannoma and pituitary adenomas Feature
VS only (N = 9857)
PA only (N = 26,546)
Both tumors (N = 31) VS size
Both tumors (N = 31) PA size
Tumor size 2303 (23.4 %)
6512 (24.5 %)
4 (12.9 %)
7 (22.6 %)
0 to \ 1 cm
Missing
1773 (18.0 %)
5498 (20.7 %)
8 (25.8 %)
6 (19.4 %)
1 to \ 2 cm
2844 (28.9 %)
5503 (20.7 %)
11 (35.5 %)
8 (25.8 %)
2 to \ 3 cm
1544 (15.7 %)
5088 (19.2 %)
5 (16.1 %)
6 (19.4 %)
3 to \ 4 cm
810 (8.2 %)
2410 (9.1 %)
2 (6.5 %)
4 (12.9 %)
C 4 cm
583 (5.9 %)
1535 (5.8 %)
1 (3.2 %)
0 (0.0 %)
1060 (10.8 %)
2322 (8.7 %)
0 (0.0 %)
1 (3.2 %)
2005
991 (10.1 %)
2512 (9.5 %)
1 (3.2 %)
0 (0.0 %)
2006
1116 (11.3 %)
2671 (10.1 %)
3 (9.7 %)
3 (9.7 %)
2007
1111 (11.3 %)
2818 (10.6 %)
2 (6.5 %)
1 (3.2 %)
2008
1119 (11.4 %)
2920 (11.0 %)
7 (22.6 %)
6 (19.4 %)
2009 2010
1182 (12.0 %) 1083 (11.0 %)
3209 (12.1 %) 3298 (12.4 %)
4 (12.9 %) 4 (12.9 %)
6 (19.4 %) 5 (16.1 %)
2011
1035 (10.5 %)
3434 (12.9 %)
6 (19.4 %)
4 (12.9 %)
2012
1160 (11.8 %)
3362 (12.7 %)
4 (12.9 %)
5 (16.1 %)
54.9 (14.8)
50.1 (18.4)
Year 2004
Age Mean (SD)
59.9 (13.4)
Under 50
3552 (36.0 %)
13,407 (50.5 %)
8 (25.8 %)
8 (25.8 %)
Age 50–59
2497 (25.3 %)
4461 (16.8 %)
7 (22.6 %)
7 (22.6 %)
16 (51.6 %)
8678 (32.7 %)
3808 (38.6 %)
16 (51.6 %)
Male
4723 (47.9 %)
11,866 (44.7 %)
20 (64.5 %)
20 (64.5 %)
Female
5134 (52.1 %)
14,680 (55.3 %)
11 (35.5 %)
11 (35.5 %)
Gender
Ethnicity 14,525 (54.75)
0.6281
0.7381
0.1820
0.2971
0.0642
0.0018
0.3109
0.0208
0.0727
0.0301
0.2304
0.2872
60.0 (13.6)
Age group
Over 60
VS ? PA versus VS VS ? PA versus PA only (P Value)* only (P-Value)*
Non-hispanic white
7345 (74.5 %)
19 (61.3 %)
19 (61.3 %)
Hispanic Non-hispanic black
821 (8.3 %) 403 (4.1 %)
4395 (16.6 %) 4701 (17.7 %)
6 (19.4 %) 1 (3.2 %)
6 (19.4 %) 1 (3.2 %)
Non-hispanic Asian or Pacific Islander
855 (8.7 %)
1984 (7.5 %)
3 (9.7 %)
3 (9.7 %)
Other or unknown
433 (4.4 %)
941 (3.5 %)
2 (6.5 %)
2 (6.5 %)
PA, pituitary adenoma; VS, vestibular schwannoma * P-values reported are from the Chi square test (categorical variables) and student t test (for the continuous variable age)
The high prevalence and unexpectedly strong familial association of each tumor, combined with the findings of this study, lead to the suspicion that the pathogenesis of VSs and PAs may share a common pathway. One hypothesis is that a common genetic basis underlies the development of both tumors. While at least 20 single-gene mutations have been discovered in sporadic PAs, few other than the NF2 gene have been identified in sporadic VSs [14, 16]. One case report of VS and PAin the same patient
drew attention to fibroblast growth factor receptor 1 (FGFR1) as a potential molecular basis for tumor development. An immunohistochemical study of FGFR1 in PAs demonstrated an inverse correlation between FGFR1 expression and tumor size [17]. On the contrary, however, analysis of 30 sporadic VSs demonstrated a positive correlation between FGFR1 overexpression and growth of VSs [18]. Since current descriptions of coexistent VS and PAare limited to case reports, a meaningful genetic
123
J Neurooncol Table 3 Observed and expected incidence calculations for vestibular schwannoma and pituitary adenomas Observed VS incidencea
Observed PA incidencea
Observed VS ? PA incidencea
Expected VS ? PA incidencea
Observed: expected rate ratio
88,254,554
1.201,071,165
2.6321589
0
0.00003161
0
88,929,243
1.115493584
2.8471762
0.00112449
0.00003151
35.68727
3 2
89,688,379 90,542,359
1.247653277 1.229258893
2.98143419 3.11345986
0.003344915 0.002208911
0.00003720 0.00003827
89.922 57.71538
2926
7
91,469,224
1.231015145
3.19889015
0.007652847
0.00003938
194.3392
3215
4
92,325,435
1.28458642
3.48224734
0.004332501
0.00004473
96.85358
1087
3303
4
93,158,638
1.166826849
3.54556493
0.004293751
0.00004137
103.7875
2011
1041
3438
6
93,886,873
1.108781203
3.66185377
0.00639067
0.00004060
157.3981
2012
1164
3367
4
94,620,901
1.230172179
3.55841042
0.004227396
0.00004377
96.57194
All years
9888
26,577
31
822,875,606
1.201639704
3.2297712
0.003767277
0.00003881
97.06920385
Year
# VS
# PA
# PA ? V S
2004
1060
2323
0
2005
992
2512
1
2006 2007
1119 1113
2674 2819
2008
1126
2009
1186
2010
Total SEER population
PA, pituitary adenoma; VS, vestibular schwannoma a
Per 100,000 persons
Table 4 Additional tumors identified in subgroup of 31 patients with coexisting vestibular schwannomas and pituitary adenomas Cancer
N (%)
Body of stomach
1 (3.2 %)
Upper lobe, lung
1 (3.2 %)
Bone marrow
1 (3.2 %)
Upper-outer quadrant of breast (UOQ)
1 (3.2 %)
Prostate gland; Prostate NOS
1 (3.2 %)
Cerebral Meninges
1 (3.2 %)
analysis of tumors in patients who manifest both has yet to be performed. The coprevalence data presented in this study may justify future tumor genome sequencing in a series of patients with coexistent VS and PA. An alternative hypothesis for the strong familiality and coprevalence of VSs and PAs is that a common environmental trigger or exposure may underlie the development of both. No known risk factors exist for the development of
Table 5 Histological subtypes of pituitary adenomas reported overall and within the select group of patients having coexisting vestibular schwannoma and pituitary adenoma
PAs in the absence of a known syndrome or mutation. Some studies have demonstrated associations between radiation exposure, noise exposure, and personal history of chickenpox and the development of VS [19]. The link between local radiofrequency exposure from cell phones and the development of VS was demonstrated in a subgroup of patients with greater than 10 years of cumulative cell phone use (particularly in the ipsilateral ear) in the Swedish INTERPHONE study [20]. A larger follow up study with more long-term cell phone users failed to show a positive correlation, however, this study instead demonstrated that patients frequently switched their preferred side of cell phone use long before the diagnosis of VS was made [21]. These data support the notion that cell phone use may increase the likelihood of detecting hearing loss as the presenting symptom of VS, but fail to convincingly defend the hypothesis that cell phones lead to their development. Postmortem histological studies regularly demonstrate that the true prevalence of many benign and malignant
ICD-0-3 Code: pituitary adenoma subtype 8140/0: Adenoma, NOS
Patients with VS ? PA
2193 (8.3 %)
1 (3.2 %)
8202/0: Microcystic adenoma
3 (0.0 %)
0 (0.0 %)
8260/0: Papillary adenoma, NOS
9 (0.0 %)
0 (0.0 %)
242 (0.9 %) 957 (3.6 %)
1 (3.2 %) 2 (6.5 %)
8270/0: Chromophobe adenoma 8271/0: Prolactinoma 8272/0: Pituitary adenoma, NOS 8280/0: Acidophil adenoma 8281/0: Mixed acidophil-basophil adenoma
123
All patients with PA
23,117 (87.0 %)
27 (87.1 %)
33 (0.1 %)
0 (0.0 %)
1 (0.0 %)
0 (0.0 %)
8290/0: Oxyphilic adenoma
16 (0.1 %)
0 (0.0 %)
8300/0: Basophil adenoma
6 (0.0 %)
0 (0.0 %)
Surgery Not specified
b
a
L Largeb F 28 Gorman and Hewer (1984) 4
Size was not specified, although ‘‘evidence of brain stem compression and obliteration of the IVth ventricle’’ was reported Size was not specified, although ‘‘large left acoustic neuroma complicated by hydrocephalus’’ was reported
Not specified Not specified Surgery HL, Imb
N ‘‘Macroadenoma’’ SRS None None L Vaamonde Lago et al. [13] 3
23
F
14
‘‘Nerve deafness’’
SRS and medical Medical Elevated growth hormone Elevated Prolactin N 16 SRS Imb, Tin ‘‘Severe’’ L 22 F Niu et al. [11] 2
42
Surgery Mild elevated Prolactin Y 40 Obs Imb L Amit et al. [12] 1
70
M
Largea
‘‘Slight’’
Vision changes Size (mm) Treatment Attributable symptoms Hearing loss Side Size (mm)a Publication
Age
Sex
Vestibular schwannoma data Demographic data No.
Table 6 Summary of 4 case reports describing the co-occurrence of vestibular schwannoma and pituitary adenoma
Pituitary adenoma data
Endocrine function
Treatment
J Neurooncol
tumors is much greater than what has been reported by clinical studies, suggesting that a portion of the population harbor subclinical disease [14, 22]. We have also observed that the clinical prevalence of many benign tumors, including VSs and PAs, has also increased secondary to more sensitive radiologic studies and greater patient access to advanced imaging [14, 23]. These phenomena, however, cannot account for the greater co-incidence of VS and PA observed in the current study for several reasons. Theoretically, any inherent biases or limitations of the SEER database should be cancelled since data from both patient populations were obtained from the same registry and our analysis utilized the incidence ratio of observed-to-expected. Additionally, patients catalogued in the SEER database presumably received appropriate imaging to establish an accurate diagnosis, and based on the study period from 2004 to 2012, the great majority of patients likely underwent MRI with contrast—the imaging modality of choice for diagnosis of both VS and PAs. Furthermore, only approximately one-fifth of all VS and PA reported in this series were less than 1 cm, and thus it is fair to assume that a significant portion were symptomatic and not simply small incidental tumors. The tumor characteristics for the 31 patients with both VS and PA were not statistically different from the population of patients with single primary tumors (Table 2). Lastly, as an internal control measure, we evaluated the co-occurrence of VS and another intracranial tumor to ensure our findings were not simply the result of methodological error or coincidental secondary diagnosis from head MRI. We found that the co-occurrence of GBM and VS in the SEER database was 17.5. Assuming that the development of VS and GBM are independent events, this suggests that a certain number of incidental tumors are diagnosed by virtue of obtaining a head MRI. However, the O/E identified for VS and GBM was still 5.5 times less than the O/E of VS and PA, further supporting a true association. In closing, there are several limitations inherent to study design that warrant consideration. The SEER database reports aggregate data from 20 regional US cancer registries, representing a selected sample of approximately 28 % of the US population. Therefore, the SEER population may not be entirely representative of the national population. Second, patient-and tumor-related variables collected in the SEER database are generic and are not tailored to the study of VS or PA. Thus many important disease-specific features within these populations are not reported. Last, beyond internal quality control measures, there is no method of SEER data verification—our analysis was limited by the quality and completeness of reporting by the individual registries. Despite these shortcomings, the SEER database affords an invaluable tool for analysis of national epidemiological trends that is particularly
123
J Neurooncol
useful for the study of rare conditions such as coexistent VS and PA.
Conclusions Using a large population based tumor registry, the current study demonstrates that the observed rate of co-incident VS and PA is greater than what is expected by chance alone. These data suggest that a common environmental or genetic predisposition exists for VS and PA development. Further study of this population may help elucidate the cause of tumorigenesis in a subset of patients with seemingly sporadic tumors.
Financial material and support Internal departmental funding was utilized without commercial sponsorship or support. Compliance with ethical standards Conflict of interest
None.
References 1. Carlson ML, Habermann EB, Wagie AE, Driscoll CL, Van Gompel JJ, Jacob JT, Link MJ (2015) The Changing Landscape of Vestibular Schwannoma Management in the United States-A Shift Toward Conservatism. Otolaryngol Head Neck Surg. doi:10.1177/0194599815590105 2. McDowell BD, Wallace RB, Carnahan RM, Chrischilles EA, Lynch CF, Schlechte JA (2011) Demographic differences in incidence for pituitary adenoma. Pituitary 14:23–30. doi:10.1007/ s11102-010-0253-4 3. Bikhazi NB, Slattery WH 3rd, Lalwani AK, Jackler RK, Bikhazi PH, Brackmann DE (1997) Familial occurrence of unilateral vestibular schwannoma. Laryngoscope 107:1176–1180 4. Neary WJ, Newton VE, Laoide-Kemp SN, Ramsden RT, Griffith G, Evans DG, Harris R, Strachan T (1996) A clinical, genetic and audiological study of patients and families with unilateral vestibular schwannomas. I. Clinical features of neurofibromatosis in patients with unilateral vestibular schwannomas. J Larnyngol Otol 110:634–640 5. Dworakowska D, Grossman AB (2009) The pathophysiology of pituitary adenomas. Best Pract Res Clin Endocrinol Metab 23:525–541. doi:10.1016/j.beem.2009.05.004 6. Couldwell WT, Cannon-Albright L (2010) A heritable predisposition to pituitary tumors. Pituitary 13:130–137. doi:10.1007/ s11102-009-0212-0 7. Committee on Hearing and Equilibrium guidelines for the evaluation of hearing preservation in acoustic neuroma (vestibular schwannoma) (1995) Otolaryngol Head Neck Surg 113:179–180 8. Overview of the SEER Program. (n.d.). Retrieved March 29, from http://seer.cancer.gov/about/overview.html
123
9. Selvin S (2004) Statistical analysis of epidemiology data. Oxford University Press, Oxford 10. Couldwell WT, Cannon-Albright LA (2014) Co-prevalence of other tumors in patients harboring pituitary tumors. J Neurosurg 121:1474–1477. doi:10.3171/2014.8.JNS132139 11. Niu Y, Ma L, Mao Q, Wu L, Chen J (2010) Pituitary adenoma and vestibular schwannoma: case report and review of the literature. J Postgrad Med 56:281–283. doi:10.4103/0022-3859.70940 12. Amit A, Achawal S, Dorward N (2008) Pituitary macro adenoma and vestibular schwannoma: a case report of dual intracranial pathologies. Br J Neurosurg 22:695–696. doi:10.1080/02688690 801987366 13. Vaamonde Lago P, Castro Vilas C, Soto Varela A, Frade Gonzalez C, Santos Perez S, Labella Caballero T (2001) Asymptomatic acoustic neurinoma associated with hypophyseal macroadenoma. Acta Otorrinolaringol Esp 52:705–708 14. Daly AF, Tichomirowa MA, Beckers A (2009) The epidemiology and genetics of pituitary adenomas. Best Pract Res Clin Endocrinol Metab 23:543–554. doi:10.1016/j.beem.2009.05.008 15. Daly AF, Vanbellinghen JF, Khoo SK, Jaffrain-Rea ML, Naves LA, Guitelman MA, Murat A, Emy P, Gimenez-Roqueplo AP, Tamburrano G, Raverot G, Barlier A, De Herder W, Penfornis A, Ciccarelli E, Estour B, Lecomte P, Gatta B, Chabre O, Sabate MI, Bertagna X, Garcia Basavilbaso N, Stalldecker G, Colao A, Ferolla P, Wemeau JL, Caron P, Sadoul JL, Oneto A, Archambeaud F, Calender A, Sinilnikova O, Montanana CF, Cavagnini F, Hana V, Solano A, Delettieres D, Luccio-Camelo DC, Basso A, Rohmer V, Brue T, Bours V, Teh BT, Beckers A (2007) Aryl hydrocarbon receptor-interacting protein gene mutations in familial isolated pituitary adenomas: analysis in 73 families. J Clin Endocrinol Metab 92:1891–1896. doi:10.1210/jc.20062513 16. Mohyuddin A, Neary WJ, Wallace A, Wu CL, Purcell S, Reid H, Ramsden RT, Read A, Black G, Evans DG (2002) Molecular genetic analysis of the NF2 gene in young patients with unilateral vestibular schwannomas. J Med Genet 39:315–322 17. Fukui S, Otani N, Nawashiro H, Yano A, Nomura N, Miyazawa T, Ohnuki A, Tsuzuki N, Katoh H, Ishihara S, Shima K (2002) Subcellular localization of basic fibroblast growth factor and fibroblast growth factor receptor 1 in pituitary adenomas. Brain tumor Pathol 19:23–29 18. O’Reilly BF, Kishore A, Crowther JA, Smith C (2004) Correlation of growth factor receptor expression with clinical growth in vestibular schwannomas. Otol Neurotol 25:791–796 19. Corona AP, Ferrite S, Lopes Mda S, Rego MA (2012) Risk factors associated with vestibular nerve schwannomas. Otol Neurotol 33:459–465. doi:10.1097/MAO.0b013e3182487fee 20. Lonn S, Ahlbom A, Hall P, Feychting M (2004) Mobile phone use and the risk of acoustic neuroma. Epidemiology 15:653–659 21. Pettersson D, Mathiesen T, Prochazka M, Bergenheim T, Florentzson R, Harder H, Nyberg G, Siesjo P, Feychting M (2014) Long-term mobile phone use and acoustic neuroma risk. Epidemiology 25:233–241. doi:10.1097/EDE.0000000000000058 22. Stewart TJ, Liland J, Schuknecht HF (1975) Occult schwannomas of the vestibular nerve. Arch Otolaryngol 101:91–95 23. Stangerup SE, Tos M, Thomsen J, Caye-Thomasen P (2010) True incidence of vestibular schwannoma? Neurosurgery 67:1335–1340. doi:10.1227/NEU.0b013e3181f22660 discussion 1340
