- Email: [email protected]
Incidence of Pineal Gland Cyst and Pineoblastoma in Children With Retinoblastoma During the Chemoreduction Era
Incidence of Pineal Gland Cyst and Pineoblastoma in Children With Retinoblastoma During the Chemoreduction Era APARNA RAMASUBRAMANIAN, CHRISTINA KYTASTY, ANNA T. MEADOWS, JERRY A. SHIELDS, ANN LEAHEY, AND CAROL L. SHIELDS PURPOSE: To report on the frequency of cysts and tumors of the pineal gland in patients with retinoblastoma. DESIGN: Observational retrospective case control study. METHODS: SETTING: Institutional. STUDY POPULATION: Four hundred eight patients treated for retinoblastoma from January 2000 to January 2012 at Wills Eye Institute, Philadelphia, Pennsylvania, USA. OBSERVATION PROCEDURE: Magnetic resonance imaging (MRI) features of the pineal gland were evaluated in all patients with retinoblastoma. Characteristics of patients with pineal cysts and pineoblastoma were reviewed. MAIN OUTCOME MEASURES: Comparison of frequency of pineal gland cyst and pineoblastoma in children managed with systemic chemoreduction vs other methods. RESULTS: Of 408 patients, treatment included systemic chemoreduction in 252 (62%) and nonchemoreduction methods in 156 (38%). Overall, 34 patients (8%) manifested pineal gland cyst and 4 (1%) showed pineoblastoma. Of all 408 patients, comparison (chemoreduction vs nonchemoreduction) revealed pineal cyst (20/252 vs 14/156, P [ .7) and pineoblastoma (1/252 vs 3/156, P [ .1). The pineal cyst (n [ 34) (mean diameter 4 mm) was asymptomatic (n [ 34), followed conservatively (n [ 34), and with minimal enlargement (n [ 2, 9%) but without progression to pineoblastoma. The cyst was found in 22 germline and 12 nongermline patients (P [ .15). Among the 4 patients with pineoblastoma, all had germline mutation and 2 had family history of retinoblastoma. Among all patients with family history of retinoblastoma (n [ 45), 2 (4%) developed pineoblastoma. The pineoblastoma was asymptomatic in 2 patients and symptomatic with vomiting and headache in 2 patients. The mean interval from date of retinoblastoma detection to pineal cyst was 2 months (median 2, range 0-8 months) and to pineoblastoma was 27 months (median 28, range 7-46 months). Management included aggressive chemotherapy and radiotherapy, with 2 survivors.
CONCLUSIONS:
Pineal gland cyst was incidentally detected in 8% of retinoblastoma patients, causing no symptoms, and without progression to pineoblastoma. Pineoblastoma was detected in 1% of patients and fewer patients who received systemic chemotherapy developed pineoblastoma, possibly indicating a systemic protective effect. (Am J Ophthalmol 2013;156:825–829. Ó 2013 by Elsevier Inc. All rights reserved.)
I
N THE PEDIATRIC POPULATION, THE PINEAL GLAND CAN
show alterations including cyst formation (pineal cyst) and solid tumor production (pineocytoma, pineoblastoma, germinoma, teratoma, astrocytoma, glial tumors).1 Pineal cyst is generally a coincidental finding in an asymptomatic child, as recognized by Karatza and associates in their study of pineal cysts in children with retinoblastoma.2 Pineoblastoma has historically been a highly fatal malignant tumor, found in children with the germline form of retinoblastoma, and preferentially in those with a positive family history.3,4 In the prechemoreduction era, pineoblastoma, also termed ‘‘trilateral retinoblastoma,’’ was found in 6% of bilateral retinoblastoma patients and 10% of patients with family history of retinoblastoma.5 Pineoblastoma is an important cause of mortality in retinoblastoma patients during in the first 5 years of life.5 Screening with neuroimaging can potentially improve survival if the tumor is detected early (<15 mm in size).3 However, with neuroimaging screening comes an increase in detection of coincidental pineal lesions.3 In this retrospective analysis, we explore the trends in the occurrence of pineal gland lesions in a consecutive series of children with retinoblastoma imaged during the chemoreduction era.
METHODS Accepted for publication May 21, 2013. From the Ocular Oncology Service, Wills Eye Institute, Thomas Jefferson University (A.R., C.K., J.A.S., C.L.S.); and Children’s Hospital of Philadelphia (A.T.M., A.L.), Philadelphia, Pennsylvania. Inquiries to Carol L. Shields, Ocular Oncology Service, Suite 1440, Wills Eye Institute, 840 Walnut Street, Philadelphia, PA 19107; e-mail: [email protected] 0002-9394/$36.00 http://dx.doi.org/10.1016/j.ajo.2013.05.023
Ó
2013 BY
THIS WAS A RETROSPECTIVE REVIEW OF MEDICAL RECORDS
of patients treated for retinoblastoma from January 1, 2000 to January 1, 2012 at Wills Eye Institute, Philadelphia, Pennsylvania, USA. The study was approved by the Institutional Review Board of Wills Eye Institute and was in compliance with Health Insurance Portability and
ELSEVIER INC. ALL
RIGHTS RESERVED.
825
TABLE. Pineal Gland Abnormalities in 408 Patients With Retinoblastoma: Demographic Features
Feature
Age (months), mean (median, range) Sex Male Female Race White African American Hispanic Asian Heredity Sporadic Familial Laterality Unilateral Bilateral Mutation Germline Nongermline
All Patients
Patients With Pineal Cyst
Patients With Pineoblastoma
(n ¼ 408)
(n ¼ 34)
(n ¼ 4)
21 (13, 0-91) 14 (9, 2-73)
5 (1, 0-14)
202 (49%) 206 (51%)
15 (44%) 19 (56%)
2 (50%) 2 (50%)
287 (70%) 60 (15%) 41 (10%) 20 (5%)
26 (76%) 4 (12%) 3 (9%) 1 (3%)
4 (100%) 0 (0%) 0 (0%) 0 (0%)
363 (89%) 45 (11%)
27 (79%) 7 (21%)
2 (50%) 2 (50%)
215 (53%) 193 (47%)
17 (50%) 17 (50%)
1 (25%) 3 (75%)
215 (53%) 193 (47%)
22 (65%) 12 (35%)
4 (100%) 0 (0%)
Values are either number (%) or mean (median, range).
Accountability Act (HIPAA) regulations. Patients included in this study were those who had undergone brain imaging with magnetic resonance imaging (MRI) during the course of treatment. Each patient was evaluated for age at diagnosis, race, sex, and hereditary pattern (sporadic, familial). Eyes were assessed for laterality of involvement (unilateral, bilateral) and total number of retinal tumors per eye, and classified according to the International Classification of Retinoblastoma. The MRI was evaluated by review of reports filed on the chart and specific information regarding abnormalities in the pineal region were documented. Those patients with follow-up MRI were documented for additional features. Statistical analysis was performed using x2 test, Fisher t test, and Mann-Whitney test. A P value of less than .05 was considered significant.
RESULTS
There was no other structural brain abnormality in any case. The demographic features of all patients and an analysis of those with pineal cyst vs those with pineoblastoma are represented in the Table. PINEAL CYST CHARACTERISTICS:
The pineal cyst was asymptomatic in all patients. The mean interval from diagnosis of retinoblastoma to detection of pineal cyst was 2 months (range 0-8 months). The mean size of pineal cyst was 4 mm (range 3-6 mm). The cyst was homogenous in 33 patients (97%) and heterogenous in 1 (3%). A ring of enhancement was seen in 18 cysts (53%). Follow-up imaging details were available for 23 patients and the cyst showed enlargement in 2 (9%), stable size in 17 (74%), and regression in 4 (17%). Among the 34 patients detected to have pineal cyst (Figure 1), 20 patients (59%) were treated with systemic chemoreduction and 14 patients (41%) were treated with other modalities. This difference did not attain statistical significance (P ¼ .7). The incidence of pineal cyst was not increased in patients with germline mutation (P ¼ .15). There were 21 patients with follow-up MRI and a mean of 4 MRIs were performed during the course of treatment. None of the cysts showed evidence of compression on surrounding structures. None of the patients with pineal cyst developed second malignancies or metastatic disease during the period of follow-up (mean follow-up 59 months, range 6-138 months).
PINEOBLASTOMA CHARACTERISTICS:
Of the total 408 patients, 4 patients were diagnosed to have pineoblastoma and 1 of these patients had received systemic chemotherapy prior to diagnosis of pineoblastoma (Figure 2). Fewer patients who received systemic chemotherapy developed pineoblastoma (1/252 vs 3/156; chemoreduction vs nonchemoreduction), though this number did not attain statistical significance (P ¼ .1). Among the 4 patients with pineoblastoma, 2 had a family history of retinoblastoma. The patients were asymptomatic in 2 cases and presented with headache and vomiting in 2 cases. The first patient was treated with surgery, chemotherapy, and radiation therapy and had no evidence of recurrence or metastatic disease at last follow-up date. The second and third patients were treated with chemotherapy, but it was discontinued because of poor response and they died. The fourth patient was treated with surgery, high-dose chemotherapy, and stem cell rescue with control of the tumor.
THERE WERE 408 PATIENTS WITH RETINOBLASTOMA WITH
adequate MRI information that were included in this analysis. The primary treatment modality was systemic chemoreduction in 252 patients (62%), enucleation in 143 (35%), external beam radiation therapy in 5 (1%), and plaque radiotherapy in 8 (2%). At initial MRI imaging, 34 patients (8%) were identified to have a pineal cyst and 4 patients (1%) were diagnosed to have pineoblastoma. 826
DISCUSSION PINEAL GLAND IMAGING HAS GREATLY IMPROVED SINCE
the development of high-resolution MRI. The European Retinoblastoma Imaging Collaboration (ERIC) recently
AMERICAN JOURNAL OF OPHTHALMOLOGY
OCTOBER 2013
FIGURE 1. Magnetic resonance image in a child with bilateral retinoblastoma and pineal cyst before (Left) and after contrast with gadolinium (Right), showing ring enhancement with contrast and minimal central enhancement.
a (multi-channel) headcoil that includes at least the following sequences:6 Transaxial fast spin-echo T2-weighted sequence (slice _4 mm) – to image anatomic and structural thickness, < abnormalities Transaxial or sagittal contrast-enhanced T1-weighted sequence (2D spin-echo T1-weighted with slice thick_3 mm; or 3D gradient-echo with slice thickness ness < < _1 mm) – to visualize the pineal gland, to rule out midline neuroectodermal tumors, and to detect leptomeningeal metastases Coronal and sagittal high-resolution T2-weighted sequence (slice thickness, 1.5 mm) – to characterize pineal gland abnormalities.
FIGURE 2. Magnetic resonance image with gadolinium contrast in a child with bilateral retinoblastoma and pineoblastoma, showing large, poorly defined lesion with uniform contrast enhancement.
published guidelines for imaging in patients with retinoblastoma.6 For the assessment of the brain and midline structures, they recommended imaging with VOL. 156, NO. 4
In the normal adult population, pineal gland cyst is diagnosed incidentally on MRI and occurs with a frequency of 1.3%, based on a large series involving 6023 individuals.7 It was earlier thought that the pineal cyst was rare in children less than 10 years of age, based on a study by Sawamura and associates7 (0% prevalence of pineal cyst in age <10 years) and Sener8 (0.6% prevalence). Al-Holou and associates evaluated 14 516 patients younger than 25 years who had undergone neuroimaging by MRI and found a 2% prevalence of pineal cyst.9 Pineal cysts were more common in female subjects (2.4%) compared to male (1.5%), and the peak prevalence was seen in the age group 6-12 years (2.7%).9 In the study by Al-Holou and associates, the prevalence in the age group 1-5 years was 1.6%, which is similar to the age group in our study. Some studies have suggested that pineal cysts and cystic degeneration of the pineal gland are seen in up to 10% of cases (all ages) at routine imaging and in 20%-40% of cases at autopsy.10 In our study the percentage of pineal cyst was 8% and the incidence was equal in male and female subjects. The incidence of pineal
PINEAL GLAND IMAGING IN RETINOBLASTOMA PATIENTS
827
cyst did not vary with prior treatment with systemic chemotherapy or with the presence of germline retinoblastoma. The slightly higher incidence of pineal cyst in our population could be related to the specific focus on imaging the pineal gland for evaluation of solid tumor, whereas previous studies retrospectively reviewed MRI scans for imaging of numerous brain conditions without specific pineal gland focus. In our study, 9% of patients showed enlargement of the cyst on follow-up imaging, which often raised the suspicion of pineoblastoma. Studies on the natural history of pineal cysts concur that the cyst can show slow growth over a period of time. Al-Holou and associates reported that 6% of patients with pineal cysts show growth when imaged over a 3-year period and the average age for patients with pineal cyst growth was 5.5 years.11 On MRI imaging, pineal cysts appear hypointense on T1 (55%-60% are somewhat hyperintense compared to cerebrospinal fluid), and hyperintense on T2 without full attenuation on fluid attenuation inversion recovery (FLAIR).1 Gadolinium contrast administration is important to differentiate a cyst from solid mass in the pineal region. Following gadolinium, most pineal gland cysts (60%) show enhancement of the cyst wall, providing a clear delineation of the structure with minimal central enhancement.12 Other brain and central nervous system abnormalities, such as Dandy-Walker malformation, cerebellar hypoplasia, cortical dysplasia, and agenesis of the corpus callosum, can be found in retinoblastoma patients with
13q deletion syndrome.13 Rodjan and associates evaluated brain MRI abnormalities in 168 patients with retinoblastoma and found pineoblastoma in 3% (5.5% of heritable group) and pineal cyst in 5.3% (2.2% in heritable group).13 Structural brain abnormality, namely corpus callosum agenesis and Dandy-Walker syndrome, was found in 1 patient each and both were associated with systemic 13q syndrome.13 In our study none of the patients was noted to have structural brain abnormalities. The incidence of pineoblastoma in our study was 1%, which is much less than the predicted occurrence in the published literature.3 Furthermore, fewer cases of pineoblastoma occurred in patients who were treated with systemic chemoreduction. Kivela¨ provided a meta-review and estimated that the chance of developing pineoblastoma was less than 0.5% among unilateral retinoblastoma, 5%-13% among sporadic bilateral retinoblastoma, and 5%-15% among familial bilateral retinoblastoma.3 The decreased incidence of pineoblastoma in our cohort possibly suggests a protective effect of systemic chemoreduction, as was originally suggested by Shields and associates.14,15 Others have commented that this could be related to the lack of external beam radiotherapy.16 In summary, pineal gland cyst was observed in 8% of retinoblastoma patients and with no relationship to pineoblastoma. The incidence of pineoblastoma has shown a downward trend in our practice, particularly during the chemoreduction era, suggesting a chemoprotective effect.
ALL AUTHORS HAVE COMPLETED AND SUBMITTED THE ICMJE FORM FOR DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST and none were reported. Support was provided by the Carlos G. Bianciotto Retinoblastoma Research Fund, Philadelphia, Pennsylvania (C.L.S.); the Lucille Wiedman Fund for Pediatric Eye Cancer, Philadelphia, Pennsylvania (J.A.S., C.L.S.); Lift for a Cure, Morrisdale, Pennsylvania (C.L.S.); and the Eye Tumor Research Foundation, Philadelphia, Pennsylvania (C.L.S., J.A.S.). The funders had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript. Contributions of authors: design (A.R., C.L.S., J.A.S., A.L.) and conduct of the study (A.R., C.K., C.L.S., A.T.M.); collection (A.R., C.K.), management (C.L.S., J.A.S., A.T.M., A.L.), analysis (A.R., A.T.M., C.L.S.), and interpretation of the data (A.R., J.A.S., C.L.S., A.L.); and preparation (A.R., C.K., C.L.S.), review (C.L.S., A.L., A.T.M., J.A.S.), or approval of the manuscript (A.R., C.K., A.T.M., A.L., C.L.S., J.A.S.).
REFERENCES 1. Bruce JN. Pineal tumors. In: Winn H, ed. Youman’s Neurological Surgery. Philadelphia: WB Saunders; 2004: 1011–1029. 2. Karatza EC, Shields CL, Flanders AE, Gonzalez ME, Shields JA. Pineal cyst simulating pinealoblastoma in 11 children with retinoblastoma. Arch Ophthalmol 2006;124(4): 595–597. 3. Kivela¨ T. Trilateral retinoblastoma: a meta-analysis of hereditary retinoblastoma associated with primary ectopic intracranial retinoblastoma. J Clin Oncol 1999;17(6):1829–1837. 4. De Potter P, Shields CL, Shields JA. Clinical variations of trilateral retinoblastoma: a report of 13 cases. J Pediatr Ophthalmol Strabismus 1994;31(1):26–31. 5. Blach LE, McCormick B, Abramson DH, Ellsworth RM. Trilateral retinoblastoma–incidence and outcome: a decade
828
6.
7.
8.
9.
of experience. Int J Radiat Oncol Biol Phys 1994;29(4): 729–733. de Graaf P, Go¨ricke S, Rodjan F, et al. Guidelines for imaging retinoblastoma: imaging principles and MRI standardization. Pediatr Radiol 2012;42(1):2–14. Sawamura Y, Ikeda J, Ozawa M, Minoshima Y, Saito H, Abe H. Magnetic resonance images reveal a high incidence of asymptomatic pineal cysts in young women. Neurosurgery 1995;37(1):11–15. Sener RN. The pineal gland: a comparative MR imaging study in children and adults with respect to normal anatomical variations and pineal cysts. Pediatr Radiol 1995;25(4): 245–248. Al-Holou WN, Garton HJ, Muraszko KM, Ibrahim M, Maher CO. Prevalence of pineal cysts in children and young adults. Clinical article. J Neurosurg Pediatr 2009;4(3): 230–236.
AMERICAN JOURNAL OF OPHTHALMOLOGY
OCTOBER 2013
10. Osborn AG, Preece MT. Intracranial cysts: radiologicpathologic correlation and imaging approach. Radiology 2006;239(3):650–664. 11. Al-Holou WN, Maher CO, Muraszko KM, Garton HJ. The natural history of pineal cysts in children and young adults. J Neurosurg Pediatr 2010;5(2):162–166. 12. Gaillard F, Jones J. Masses of the pineal region: clinical presentation and radiographic features. Postgrad Med J 2010; 86(1020):597–607. 13. Rodjan F, de Graaf P, Moll AC, et al. Brain abnormalities on MR imaging in patients with retinoblastoma. AJNR Am J Neuroradiol 2010;31(8):1385–1389.
VOL. 156, NO. 4
14. Shields CL, Shields JA, Meadows AT. Chemoreduction for retinoblastoma may prevent trilateral retinoblastoma (Letter). J Clin Oncol 2000;18(1):236–237. 15. Shields CL, Meadows AT, Shields JA, Carvalho C, Smith AF. Chemoreduction for retinoblastoma may prevent intracranial neuroblastic malignancy (trilateral retinoblastoma). Arch Ophthalmol 2001;119(9): 1269–1272. 16. Marees T, Moll AC, Imhof SM, de Boer MR, Ringens PJ, van Leeuwen FE. Re: More about second cancers after retinoblastoma. J Natl Cancer Inst 2010;102(11): 831–832.
PINEAL GLAND IMAGING IN RETINOBLASTOMA PATIENTS
829
Biosketch Dr Aparna Ramasubramanian graduated from Medical College Trivandrum, India and then finished ophthalmology residency at PGIMER, Chandigarh, India. She then did ocular oncology fellowhip at Wills Eye Institute, Philadelphia and pediatric ophthalmology fellowship at Boston Childrens Hospital. Her career interest is retinoblastoma and she has been keenly involved in research on retinoblastoma which have been published in many peer- reviewed journals. She has also co-authored a book on retinoblastoma.
829.e1
AMERICAN JOURNAL OF OPHTHALMOLOGY
OCTOBER 2013
Biosketch Dr Carol L. Shields completed her residency in ophthalmology at Wills Eye Hospital in Philadelphia and subsequently did fellowship training in ocular oncology, oculoplastic surgery, and ophthalmic pathology. She is currently Co-Director of the Oncology Service, Wills Eye Hospital, and Professor of Ophthalmology at Thomas Jefferson University in Philadelphia. She has authored or coauthored 7 textbooks, over 1000 articles in major journals, 300 textbook chapters, given nearly 600 lectureships, and has received numerous professional awards.
VOL. 156, NO. 4
PINEAL GLAND IMAGING IN RETINOBLASTOMA PATIENTS
829.e2
CONCLUSIONS:
Pineal gland cyst was incidentally detected in 8% of retinoblastoma patients, causing no symptoms, and without progression to pineoblastoma. Pineoblastoma was detected in 1% of patients and fewer patients who received systemic chemotherapy developed pineoblastoma, possibly indicating a systemic protective effect. (Am J Ophthalmol 2013;156:825–829. Ó 2013 by Elsevier Inc. All rights reserved.)
I
N THE PEDIATRIC POPULATION, THE PINEAL GLAND CAN
show alterations including cyst formation (pineal cyst) and solid tumor production (pineocytoma, pineoblastoma, germinoma, teratoma, astrocytoma, glial tumors).1 Pineal cyst is generally a coincidental finding in an asymptomatic child, as recognized by Karatza and associates in their study of pineal cysts in children with retinoblastoma.2 Pineoblastoma has historically been a highly fatal malignant tumor, found in children with the germline form of retinoblastoma, and preferentially in those with a positive family history.3,4 In the prechemoreduction era, pineoblastoma, also termed ‘‘trilateral retinoblastoma,’’ was found in 6% of bilateral retinoblastoma patients and 10% of patients with family history of retinoblastoma.5 Pineoblastoma is an important cause of mortality in retinoblastoma patients during in the first 5 years of life.5 Screening with neuroimaging can potentially improve survival if the tumor is detected early (<15 mm in size).3 However, with neuroimaging screening comes an increase in detection of coincidental pineal lesions.3 In this retrospective analysis, we explore the trends in the occurrence of pineal gland lesions in a consecutive series of children with retinoblastoma imaged during the chemoreduction era.
METHODS Accepted for publication May 21, 2013. From the Ocular Oncology Service, Wills Eye Institute, Thomas Jefferson University (A.R., C.K., J.A.S., C.L.S.); and Children’s Hospital of Philadelphia (A.T.M., A.L.), Philadelphia, Pennsylvania. Inquiries to Carol L. Shields, Ocular Oncology Service, Suite 1440, Wills Eye Institute, 840 Walnut Street, Philadelphia, PA 19107; e-mail: [email protected] 0002-9394/$36.00 http://dx.doi.org/10.1016/j.ajo.2013.05.023
Ó
2013 BY
THIS WAS A RETROSPECTIVE REVIEW OF MEDICAL RECORDS
of patients treated for retinoblastoma from January 1, 2000 to January 1, 2012 at Wills Eye Institute, Philadelphia, Pennsylvania, USA. The study was approved by the Institutional Review Board of Wills Eye Institute and was in compliance with Health Insurance Portability and
ELSEVIER INC. ALL
RIGHTS RESERVED.
825
TABLE. Pineal Gland Abnormalities in 408 Patients With Retinoblastoma: Demographic Features
Feature
Age (months), mean (median, range) Sex Male Female Race White African American Hispanic Asian Heredity Sporadic Familial Laterality Unilateral Bilateral Mutation Germline Nongermline
All Patients
Patients With Pineal Cyst
Patients With Pineoblastoma
(n ¼ 408)
(n ¼ 34)
(n ¼ 4)
21 (13, 0-91) 14 (9, 2-73)
5 (1, 0-14)
202 (49%) 206 (51%)
15 (44%) 19 (56%)
2 (50%) 2 (50%)
287 (70%) 60 (15%) 41 (10%) 20 (5%)
26 (76%) 4 (12%) 3 (9%) 1 (3%)
4 (100%) 0 (0%) 0 (0%) 0 (0%)
363 (89%) 45 (11%)
27 (79%) 7 (21%)
2 (50%) 2 (50%)
215 (53%) 193 (47%)
17 (50%) 17 (50%)
1 (25%) 3 (75%)
215 (53%) 193 (47%)
22 (65%) 12 (35%)
4 (100%) 0 (0%)
Values are either number (%) or mean (median, range).
Accountability Act (HIPAA) regulations. Patients included in this study were those who had undergone brain imaging with magnetic resonance imaging (MRI) during the course of treatment. Each patient was evaluated for age at diagnosis, race, sex, and hereditary pattern (sporadic, familial). Eyes were assessed for laterality of involvement (unilateral, bilateral) and total number of retinal tumors per eye, and classified according to the International Classification of Retinoblastoma. The MRI was evaluated by review of reports filed on the chart and specific information regarding abnormalities in the pineal region were documented. Those patients with follow-up MRI were documented for additional features. Statistical analysis was performed using x2 test, Fisher t test, and Mann-Whitney test. A P value of less than .05 was considered significant.
RESULTS
There was no other structural brain abnormality in any case. The demographic features of all patients and an analysis of those with pineal cyst vs those with pineoblastoma are represented in the Table. PINEAL CYST CHARACTERISTICS:
The pineal cyst was asymptomatic in all patients. The mean interval from diagnosis of retinoblastoma to detection of pineal cyst was 2 months (range 0-8 months). The mean size of pineal cyst was 4 mm (range 3-6 mm). The cyst was homogenous in 33 patients (97%) and heterogenous in 1 (3%). A ring of enhancement was seen in 18 cysts (53%). Follow-up imaging details were available for 23 patients and the cyst showed enlargement in 2 (9%), stable size in 17 (74%), and regression in 4 (17%). Among the 34 patients detected to have pineal cyst (Figure 1), 20 patients (59%) were treated with systemic chemoreduction and 14 patients (41%) were treated with other modalities. This difference did not attain statistical significance (P ¼ .7). The incidence of pineal cyst was not increased in patients with germline mutation (P ¼ .15). There were 21 patients with follow-up MRI and a mean of 4 MRIs were performed during the course of treatment. None of the cysts showed evidence of compression on surrounding structures. None of the patients with pineal cyst developed second malignancies or metastatic disease during the period of follow-up (mean follow-up 59 months, range 6-138 months).
PINEOBLASTOMA CHARACTERISTICS:
Of the total 408 patients, 4 patients were diagnosed to have pineoblastoma and 1 of these patients had received systemic chemotherapy prior to diagnosis of pineoblastoma (Figure 2). Fewer patients who received systemic chemotherapy developed pineoblastoma (1/252 vs 3/156; chemoreduction vs nonchemoreduction), though this number did not attain statistical significance (P ¼ .1). Among the 4 patients with pineoblastoma, 2 had a family history of retinoblastoma. The patients were asymptomatic in 2 cases and presented with headache and vomiting in 2 cases. The first patient was treated with surgery, chemotherapy, and radiation therapy and had no evidence of recurrence or metastatic disease at last follow-up date. The second and third patients were treated with chemotherapy, but it was discontinued because of poor response and they died. The fourth patient was treated with surgery, high-dose chemotherapy, and stem cell rescue with control of the tumor.
THERE WERE 408 PATIENTS WITH RETINOBLASTOMA WITH
adequate MRI information that were included in this analysis. The primary treatment modality was systemic chemoreduction in 252 patients (62%), enucleation in 143 (35%), external beam radiation therapy in 5 (1%), and plaque radiotherapy in 8 (2%). At initial MRI imaging, 34 patients (8%) were identified to have a pineal cyst and 4 patients (1%) were diagnosed to have pineoblastoma. 826
DISCUSSION PINEAL GLAND IMAGING HAS GREATLY IMPROVED SINCE
the development of high-resolution MRI. The European Retinoblastoma Imaging Collaboration (ERIC) recently
AMERICAN JOURNAL OF OPHTHALMOLOGY
OCTOBER 2013
FIGURE 1. Magnetic resonance image in a child with bilateral retinoblastoma and pineal cyst before (Left) and after contrast with gadolinium (Right), showing ring enhancement with contrast and minimal central enhancement.
a (multi-channel) headcoil that includes at least the following sequences:6 Transaxial fast spin-echo T2-weighted sequence (slice _4 mm) – to image anatomic and structural thickness, < abnormalities Transaxial or sagittal contrast-enhanced T1-weighted sequence (2D spin-echo T1-weighted with slice thick_3 mm; or 3D gradient-echo with slice thickness ness < < _1 mm) – to visualize the pineal gland, to rule out midline neuroectodermal tumors, and to detect leptomeningeal metastases Coronal and sagittal high-resolution T2-weighted sequence (slice thickness, 1.5 mm) – to characterize pineal gland abnormalities.
FIGURE 2. Magnetic resonance image with gadolinium contrast in a child with bilateral retinoblastoma and pineoblastoma, showing large, poorly defined lesion with uniform contrast enhancement.
published guidelines for imaging in patients with retinoblastoma.6 For the assessment of the brain and midline structures, they recommended imaging with VOL. 156, NO. 4
In the normal adult population, pineal gland cyst is diagnosed incidentally on MRI and occurs with a frequency of 1.3%, based on a large series involving 6023 individuals.7 It was earlier thought that the pineal cyst was rare in children less than 10 years of age, based on a study by Sawamura and associates7 (0% prevalence of pineal cyst in age <10 years) and Sener8 (0.6% prevalence). Al-Holou and associates evaluated 14 516 patients younger than 25 years who had undergone neuroimaging by MRI and found a 2% prevalence of pineal cyst.9 Pineal cysts were more common in female subjects (2.4%) compared to male (1.5%), and the peak prevalence was seen in the age group 6-12 years (2.7%).9 In the study by Al-Holou and associates, the prevalence in the age group 1-5 years was 1.6%, which is similar to the age group in our study. Some studies have suggested that pineal cysts and cystic degeneration of the pineal gland are seen in up to 10% of cases (all ages) at routine imaging and in 20%-40% of cases at autopsy.10 In our study the percentage of pineal cyst was 8% and the incidence was equal in male and female subjects. The incidence of pineal
PINEAL GLAND IMAGING IN RETINOBLASTOMA PATIENTS
827
cyst did not vary with prior treatment with systemic chemotherapy or with the presence of germline retinoblastoma. The slightly higher incidence of pineal cyst in our population could be related to the specific focus on imaging the pineal gland for evaluation of solid tumor, whereas previous studies retrospectively reviewed MRI scans for imaging of numerous brain conditions without specific pineal gland focus. In our study, 9% of patients showed enlargement of the cyst on follow-up imaging, which often raised the suspicion of pineoblastoma. Studies on the natural history of pineal cysts concur that the cyst can show slow growth over a period of time. Al-Holou and associates reported that 6% of patients with pineal cysts show growth when imaged over a 3-year period and the average age for patients with pineal cyst growth was 5.5 years.11 On MRI imaging, pineal cysts appear hypointense on T1 (55%-60% are somewhat hyperintense compared to cerebrospinal fluid), and hyperintense on T2 without full attenuation on fluid attenuation inversion recovery (FLAIR).1 Gadolinium contrast administration is important to differentiate a cyst from solid mass in the pineal region. Following gadolinium, most pineal gland cysts (60%) show enhancement of the cyst wall, providing a clear delineation of the structure with minimal central enhancement.12 Other brain and central nervous system abnormalities, such as Dandy-Walker malformation, cerebellar hypoplasia, cortical dysplasia, and agenesis of the corpus callosum, can be found in retinoblastoma patients with
13q deletion syndrome.13 Rodjan and associates evaluated brain MRI abnormalities in 168 patients with retinoblastoma and found pineoblastoma in 3% (5.5% of heritable group) and pineal cyst in 5.3% (2.2% in heritable group).13 Structural brain abnormality, namely corpus callosum agenesis and Dandy-Walker syndrome, was found in 1 patient each and both were associated with systemic 13q syndrome.13 In our study none of the patients was noted to have structural brain abnormalities. The incidence of pineoblastoma in our study was 1%, which is much less than the predicted occurrence in the published literature.3 Furthermore, fewer cases of pineoblastoma occurred in patients who were treated with systemic chemoreduction. Kivela¨ provided a meta-review and estimated that the chance of developing pineoblastoma was less than 0.5% among unilateral retinoblastoma, 5%-13% among sporadic bilateral retinoblastoma, and 5%-15% among familial bilateral retinoblastoma.3 The decreased incidence of pineoblastoma in our cohort possibly suggests a protective effect of systemic chemoreduction, as was originally suggested by Shields and associates.14,15 Others have commented that this could be related to the lack of external beam radiotherapy.16 In summary, pineal gland cyst was observed in 8% of retinoblastoma patients and with no relationship to pineoblastoma. The incidence of pineoblastoma has shown a downward trend in our practice, particularly during the chemoreduction era, suggesting a chemoprotective effect.
ALL AUTHORS HAVE COMPLETED AND SUBMITTED THE ICMJE FORM FOR DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST and none were reported. Support was provided by the Carlos G. Bianciotto Retinoblastoma Research Fund, Philadelphia, Pennsylvania (C.L.S.); the Lucille Wiedman Fund for Pediatric Eye Cancer, Philadelphia, Pennsylvania (J.A.S., C.L.S.); Lift for a Cure, Morrisdale, Pennsylvania (C.L.S.); and the Eye Tumor Research Foundation, Philadelphia, Pennsylvania (C.L.S., J.A.S.). The funders had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript. Contributions of authors: design (A.R., C.L.S., J.A.S., A.L.) and conduct of the study (A.R., C.K., C.L.S., A.T.M.); collection (A.R., C.K.), management (C.L.S., J.A.S., A.T.M., A.L.), analysis (A.R., A.T.M., C.L.S.), and interpretation of the data (A.R., J.A.S., C.L.S., A.L.); and preparation (A.R., C.K., C.L.S.), review (C.L.S., A.L., A.T.M., J.A.S.), or approval of the manuscript (A.R., C.K., A.T.M., A.L., C.L.S., J.A.S.).
REFERENCES 1. Bruce JN. Pineal tumors. In: Winn H, ed. Youman’s Neurological Surgery. Philadelphia: WB Saunders; 2004: 1011–1029. 2. Karatza EC, Shields CL, Flanders AE, Gonzalez ME, Shields JA. Pineal cyst simulating pinealoblastoma in 11 children with retinoblastoma. Arch Ophthalmol 2006;124(4): 595–597. 3. Kivela¨ T. Trilateral retinoblastoma: a meta-analysis of hereditary retinoblastoma associated with primary ectopic intracranial retinoblastoma. J Clin Oncol 1999;17(6):1829–1837. 4. De Potter P, Shields CL, Shields JA. Clinical variations of trilateral retinoblastoma: a report of 13 cases. J Pediatr Ophthalmol Strabismus 1994;31(1):26–31. 5. Blach LE, McCormick B, Abramson DH, Ellsworth RM. Trilateral retinoblastoma–incidence and outcome: a decade
828
6.
7.
8.
9.
of experience. Int J Radiat Oncol Biol Phys 1994;29(4): 729–733. de Graaf P, Go¨ricke S, Rodjan F, et al. Guidelines for imaging retinoblastoma: imaging principles and MRI standardization. Pediatr Radiol 2012;42(1):2–14. Sawamura Y, Ikeda J, Ozawa M, Minoshima Y, Saito H, Abe H. Magnetic resonance images reveal a high incidence of asymptomatic pineal cysts in young women. Neurosurgery 1995;37(1):11–15. Sener RN. The pineal gland: a comparative MR imaging study in children and adults with respect to normal anatomical variations and pineal cysts. Pediatr Radiol 1995;25(4): 245–248. Al-Holou WN, Garton HJ, Muraszko KM, Ibrahim M, Maher CO. Prevalence of pineal cysts in children and young adults. Clinical article. J Neurosurg Pediatr 2009;4(3): 230–236.
AMERICAN JOURNAL OF OPHTHALMOLOGY
OCTOBER 2013
10. Osborn AG, Preece MT. Intracranial cysts: radiologicpathologic correlation and imaging approach. Radiology 2006;239(3):650–664. 11. Al-Holou WN, Maher CO, Muraszko KM, Garton HJ. The natural history of pineal cysts in children and young adults. J Neurosurg Pediatr 2010;5(2):162–166. 12. Gaillard F, Jones J. Masses of the pineal region: clinical presentation and radiographic features. Postgrad Med J 2010; 86(1020):597–607. 13. Rodjan F, de Graaf P, Moll AC, et al. Brain abnormalities on MR imaging in patients with retinoblastoma. AJNR Am J Neuroradiol 2010;31(8):1385–1389.
VOL. 156, NO. 4
14. Shields CL, Shields JA, Meadows AT. Chemoreduction for retinoblastoma may prevent trilateral retinoblastoma (Letter). J Clin Oncol 2000;18(1):236–237. 15. Shields CL, Meadows AT, Shields JA, Carvalho C, Smith AF. Chemoreduction for retinoblastoma may prevent intracranial neuroblastic malignancy (trilateral retinoblastoma). Arch Ophthalmol 2001;119(9): 1269–1272. 16. Marees T, Moll AC, Imhof SM, de Boer MR, Ringens PJ, van Leeuwen FE. Re: More about second cancers after retinoblastoma. J Natl Cancer Inst 2010;102(11): 831–832.
PINEAL GLAND IMAGING IN RETINOBLASTOMA PATIENTS
829
Biosketch Dr Aparna Ramasubramanian graduated from Medical College Trivandrum, India and then finished ophthalmology residency at PGIMER, Chandigarh, India. She then did ocular oncology fellowhip at Wills Eye Institute, Philadelphia and pediatric ophthalmology fellowship at Boston Childrens Hospital. Her career interest is retinoblastoma and she has been keenly involved in research on retinoblastoma which have been published in many peer- reviewed journals. She has also co-authored a book on retinoblastoma.
829.e1
AMERICAN JOURNAL OF OPHTHALMOLOGY
OCTOBER 2013
Biosketch Dr Carol L. Shields completed her residency in ophthalmology at Wills Eye Hospital in Philadelphia and subsequently did fellowship training in ocular oncology, oculoplastic surgery, and ophthalmic pathology. She is currently Co-Director of the Oncology Service, Wills Eye Hospital, and Professor of Ophthalmology at Thomas Jefferson University in Philadelphia. She has authored or coauthored 7 textbooks, over 1000 articles in major journals, 300 textbook chapters, given nearly 600 lectureships, and has received numerous professional awards.
VOL. 156, NO. 4
PINEAL GLAND IMAGING IN RETINOBLASTOMA PATIENTS
829.e2