- Email: [email protected]
Metastatic Retinoblastoma
Metastatic Retinoblastoma Clinical Features, Treatment, and Prognosis Kaan Gündüz, MD,1 Orkun Müftüoglu, MD,1 I˙lhan Günalp, MD,1 Emel Ünal, MD,2 Nurdan Taçyıldız, MD2 Purpose: To evaluate the clinical features, treatment, and prognosis in patients with metastatic retinoblastoma. Design: Retrospective interventional case series. Participants: Eighteen consecutive patients with metastatic retinoblastoma who were diagnosed and managed at the Oncology Service of Ankara University, Turkey, between January 1999 and January 2005. Methods: All patients underwent magnetic resonance imaging (MRI) of the orbit and brain, lumbar puncture, bone marrow aspiration, and bone scintigraphy for metastatic evaluation. Histopathologic confirmation of retinoblastoma via enucleation, exenteration, or orbital biopsy was obtained in each patient. The status of extraocular spread (optic nerve, extrascleral extension, or both) was assessed based on histopathologic or MRI results. Systemic treatment for metastatic retinoblastoma consisted of chemotherapy and radiotherapy (craniospinal, orbital, or both), if necessary. Main Outcome Measures: Status of extraocular spread, site of metastasis, and survival from metastatic retinoblastoma. Results: At presentation, the mean patient age was 45 months (range, 13– 86). Ten patients had unilateral retinoblastoma, 7 had bilateral retinoblastoma, and 1 had trilateral retinoblastoma. All patients with metastatic retinoblastoma had histopathologic or MRI evidence of unilateral extraocular disease characterized by optic nerve involvement, extrascleral extension, or both. Nine of 18 patients experienced central nervous system (CNS) involvement, 5 patients had distant and CNS metastasis, and 4 patients had distant metastasis only. Fourteen patients underwent craniospinal irradiation and 12 had orbital irradiation. At a mean follow-up of 24 months (range, 4 – 62), all patients with CNS and concurrent distant and CNS metastasis were deceased. Four patients who had distant metastasis only were alive at a follow-up ranging from 9 to 62 months. Conclusions: The prognosis for metastatic retinoblastoma is dismal and the presence of CNS involvement may portend an even worse outcome. Ophthalmology 2006;113:1558 –1566 © 2006 by the American Academy of Ophthalmology.
Retinoblastoma is the most common intraocular malignant tumor in children. In neglected or untreated cases, retinoblastoma can demonstrate extraocular spread, primarily through the optic nerve,1 but also through the sclera.2 Orbital invasion via either route is a risk factor for metastasis.3 Retinoblastoma can metastasize to the regional lymph nodes, central nervous system (CNS), and distant organs, usually bone and bone marrow.3,4 In contrast to patients with intraocular retinoblastoma, patients with metastatic retinoblastoma have a poor prognosis.3–7 Metastatic retinoblastoma is rare in advanced countries. However, it is a major cause of retinoblastoma-related mortality in developing countries. In 2 reports from developing countries, the frequency of metastatic retinoblastoma was quoted as approximately 9%8 and 11%.9 Originally received: September 22, 2005. Accepted: March 27, 2006. Manuscript no. 2005-853. 1 Department of Ophthalmology, Ankara University Faculty of Medicine, Ankara, Turkey. 2 Department of Pediatric Oncology, Ankara University Faculty of Medicine, Ankara, Turkey. Correspondence to Kaan Gündüz, MD, Mesa Koza Plaza 24/18, Küpe Sok, GOP 06700, Ankara, Turkey. E-mail: [email protected].
1558
© 2006 by the American Academy of Ophthalmology Published by Elsevier Inc.
The purpose of our study was to evaluate the clinical features, treatment, and prognosis of patients with metastatic retinoblastoma diagnosed and treated at the Ocular and Pediatric Oncology Services, Ankara University, Ankara, Turkey.
Patients and Methods The study was a retrospective, interventional case series of 18 consecutive patients with metastatic retinoblastoma who were diagnosed and treated at our oncology services between January 1999 and January 2005. Institutional review board approval was obtained for the various surgical procedures and other treatments given. The patients’ families were informed about the expected natural course of the disease, the surgical procedures, and other treatments involved, and written consent was obtained from the parents of each patient. Data recorded included age at diagnosis, gender, hereditary pattern (familial, sporadic), and laterality (unilateral, bilateral). Each patient underwent complete ocular examination under anesthesia. Intraocular pressure and anterior segment findings, including iris neovascularization, pseudohypopyon, and funduscopic features, were noted. Intraocular tumor dimensions and Reese-Ellsworth groups of retinoblastoma were noted for each eye, if applicable (i.e., if the tumor was intraocular). Patients with intraocular tumors also underwent A- and B-mode ultrasonograISSN 0161-6420/06/$–see front matter doi:10.1016/j.ophtha.2006.03.039
Gündüz et al 䡠 Metastatic Retinoblastoma phy. All patients underwent orbital and cranial magnetic resonance imaging (MRI). Each patient had histopathologic confirmation of retinoblastoma established via exenteration, enucleation of either the currently affected or the previously affected fellow eye, or orbital biopsy. Patients with massive orbital involvement had orbital exenteration. Patients with advanced intraocular disease were treated with enucleation. In some cases, extraocular spread of retinoblastoma resulted in an orbital mass. These patients underwent an orbital biopsy procedure. Based on the histopathologic and MRI data, the status of extraocular spread (retrolaminar optic nerve involvement, tumor at cut-end of optic nerve, orbital optic nerve involvement, extrascleral involvement, or a combination thereof) was assessed. As soon as extraocular spread was determined, all patients underwent lumbar puncture, bone marrow aspiration, and bone scintigraphy for metastatic evaluation. After the diagnosis of metastatic retinoblastoma was made, systemic treatment was instituted. Systemic treatment consisted of chemotherapy (regimen A or B)10,11 and craniospinal external-beam radiotherapy (24 Gy to the cranium and 18 Gy to the spine), if necessary (Table 1). The first 4 patients enrolled in this study had regimen A chemotherapy and the following 14 had regimen B chemotherapy. Patients with orbital disease (retinoblastoma infiltrating the cut end of the optic nerve, orbital optic nerve, massive extrascleral spread, or a combination thereof) received orbital external-beam radiotherapy (EBRT) at a dose of 45 Gy. Some changes were made in the original chemotherapy regimens. In regimen A, the dose of cyclophosphamide was reduced to 20 mg/kg and cisplatin was replaced with carboplatin, the latter having less nephrotoxicity and ototoxicity.10 Patients were followed up at regular chemotherapy intervals and at 2- to 3-month intervals after the cessation of chemotherapy.
Table 1. Chemotherapy Regimens Used in Metastatic Retinoblastoma Regimen A Weeks 0, 6, 12, 18, 24, 27, 30, 33 Cyclophosphamide (20 mg/kg, day 1) Doxorubicin (0.67 mg/kg, days 1–3) Vincristine (0.05 mg/kg, day 1) Weeks 3, 9, 15, 21 Carboplatin (560 mg/m2, day 1) Etoposide (3.3 mg/kg, days 1–3) Weeks 36, 39, 42, 45, 48, 51, 54, 57 Cyclophosphamide (30 mg/kg, day 1) Vincristine (0.05 mg/kg, day 1) Intrathecal chemotherapy* (weeks 0, 1, 2, 3, 4, 5, 6, and at every 3-week chemotherapy cycle after week 6) Age ⱕ2 years: methotrexate (6.0 mg), cytosine arabinoside (12.0 mg), dexamethasone (2.0 mg/m2) Age ⬎2–3 years: methotrexate (9.0 mg), cytosine arabinoside (18.0 mg), dexamethasone (2.0 mg/m2) Age ⬎3–9 years: methotrexate (12.0 mg), cytosine arabinoside (24.0 mg), dexamethasone (2.0 mg/m2) Age ⬎9 years: methotrexate (12.0 mg), cytosine arabinoside (30.0 mg), dexamethasone (2.0 mg/m2) Regimen B (ICE ⫽ ifosfamide, carboplatin, etoposide) Weeks 1, 4, 7, 10, 13, 16 Ifosfamide (1.8 g/m2/day, days 1–5) Carboplatin (560 mg/m2/day, day 1) Etoposide (100 mg/m2/day, days 1–5) (No intrathecal chemotherapy) *In patients with central nervous system involvement.
Results Eighteen consecutive patients with metastatic retinoblastoma were included in this study. Patient demographics, ocular status, and surgical methods used are shown in Table 2. Mean patient age was 45 months (range, 13– 86 months). Ten patients (55%) were males and 8 (45%) were females. The disease was sporadic in 17 patients (94%), and 1 patient (6%) had a family history of retinoblastoma. Retinoblastoma was bilateral in 7 (38%) patients, unilateral in 10 (56%) patients, and trilateral in 1 (6%) patient. Patients with metastatic retinoblastoma were divided into 3 groups according to the ocular presentation patterns described below. Group 1 included 6 eyes of 6 patients (patients 1– 6) with clinically manifest extraocular retinoblastoma at presentation. The diagnosis of extraocular retinoblastoma was made clinically and by neuroimaging studies. The mean age at presentation was 62 months (range, 26 – 86 months). The delay in presentation of these patients was the result of physician misdiagnosis, lack of health insurance, or both, which prevented the families from seeking proper medical care. One eye had intraocular recurrence of retinoblastoma 46 months after EBRT, and the tumor subsequently spread extraocularly (patient 2). Three patients in group 1 underwent lid-sparing exenteration because of massive orbital tumor (Fig 1), and 2 patients underwent orbital biopsy to establish histopathologic diagnosis. One patient (patient 4) underwent enucleation of the fellow eye 2 years previously, and the diagnosis of retinoblastoma was established histopathologically (Fig 2A). Surgery was not carried out in the other eye of this patient. Group 2 included 9 eyes of 9 patients (patients 7–15) who had advanced intraocular disease on clinical examination and had histopathologic evidence of extraocular spread on enucleation. The mean age at presentation was 36 months (range,13–70 months). Two of 9 eyes (patients 8 and 11) had diffuse infiltrating retinoblastoma and 7 of 9 eyes (patients 7, 9, 10, 13, and 14) had Reese-Ellsworth group 5B retinoblastoma in the affected eyes. The optic disc was not visible in the affected eyes of 5 of these 7 patients, because the ocular media was hazy or the tumor was overhanging the optic nerve. The remaining 2 of 7 eyes (patients 12 and 15) had massive recurrent intraocular retinoblastoma. One patient (patient 15) had been treated with primary EBRT 10 years previously at our oncology service and had been followed up for 37 months. The patient then was lost to follow-up until the tumor recurred. The other patient (patient 12) had received chemotherapy (carboplatin) elsewhere. All 9 eyes in group 2 underwent enucleation. Group 3 was composed of 3 eyes of 3 patients who had undergone surgical intervention (evisceration, patient 16; pars plana vitrectomy, patient 17; anterior chamber washout, patient 18) elsewhere because of misdiagnosis of an unsuspected retinoblastoma (Fig 3). Patient 17 underwent pars plana vitrectomy for a retinal lesion diagnosed to be a hamartoma. The patient developed extrascleral extension at the vitrectomy port and experienced subsequent orbital involvement. The mean age at presentation to our oncology service was 39 months (range, 26 –56 months). All 3 eyes were enucleated at our service. Table 3 shows the extraocular spread status, metastatic sites, radiotherapy and chemotherapy administrations, follow-up period, and final patient status. Based on the histopathologic and MRI data, all 18 patients with metastatic retinoblastoma had extraocular extension. Five of 18 patients (patients 7, 8, 11, 16, and 17) had retrolaminar optic nerve involvement, 4 (patients 9, 10, 15, and 18) had tumor at the optic nerve cut end, 6 (patients 1– 6) had tumor in the orbital optic nerve, and 17 (all patients except patient 17) had extrascleral extension. Extraocular extension occurred unilaterally in every case. Twelve of 18 patients (patients 1– 6, 8 –11, 15, 18) had massive orbital retinoblastoma and received orbital radiotherapy (45 Gy).
1559
Ophthalmology Volume 113, Number 9, September 2006 Table 2. Demographics and Examination Results of Metastatic Retinoblastoma Patients at Presentation Age at Patient Diagnosis No. Gender (yrs) Laterality
Ocular Status
Surgery
Group 1 (manifest extraocular orbital disease) 1 M 4.2 B Buphthalmos, orbital mass, massive proptosis 2 F 4.8 B Enlarged and perforated globe, orbital mass, massive proptosis (recurrence 46 mos after EBRT) 3 M 5.8 U Buphthalmos, hemorrhagic orbital mass 4 F 7.2 B Buphthalmos, orbital mass 5 F 2.2 T Buphthalmos, orbital mass 6 M 7.0 U Buphthalmos, orbital mass diagnosed as endophthalmitis elsewhere
EXE EXE EXE E (fellow eye) Biopsy Biopsy
Anterior Segment Findings Tumor Dimension (mm) Group 2 (advanced intraocular disease) 7 F 5.8 U
Pseudohypyon
Iris Neovascularization
Neovascular Glaucoma
Reese–Ellsworth Classification
–
–
–
5B
E
– – – ⫹ – – – –
– ⫹ ⫹ ⫹ – – – –
– – ⫹ ⫹
N/A 5B 5B N/A 5B 5B 5B 5B
E E E E E E E E
5.7 2.5 2.2 2.3 1.6 1.1 2.8 2.7
U B U U B U U B
Tumor filling vitreous cavity DIR 30 ⫻ 25 ⫻ 12 30 ⫻ 20 ⫻ 7 DIR 7.5 ⫻ 6.5 ⫻ 3 15 ⫻ 10 ⫻ 8 6⫻6⫻2 25 ⫻ 15 ⫻ 8
Group 3 (misdiagnosis) 16 M 4.7 17 M 2.9 18 M 2.2
U U B
S/P Evisceration due to non-resolving VH (elsewhere) S/P PPV for presumed hamartoma (elsewhere) S/P Anterior chamber tap for non-resolving hyphema (elsewhere)
8 9 10 11 12 13 14 15
M F F M F F M M
E E E
B ⫽ bilateral; DIR ⫽ diffuse infiltrating retinoblastoma; E ⫽ enucleation; EBRT ⫽ external beam radiotherapy; EXE ⫽ exenteration; F ⫽ female; M ⫽ male; N/A ⫽ not available; PPV ⫽ pars plana vitrectomy; Pseudohpy. ⫽ pseudohypopyon; S/P ⫽ status post; T ⫽ trilateral; U ⫽ unilateral; VH ⫽ vitreous hemorrhage. ⫹, present; ⫺, absent.
Systemic evaluation showed that 9 of 18 patients (patients 1, 3– 6, 7, 9, 10, and 14) had CNS metastasis, 5 (patients 2, 8, 11, 13, and 16) had concurrent CNS and distant organ metastasis, and 4 (patients 12, 15, 17, and 18) had distant organ metastasis. All patients underwent systemic chemotherapy (regimen A or B; Fig 2). The major complication of systemic chemotherapy was myelosuppression and occasional episodes of neutropenic fever, which occurred in all patients and was treated appropriately. Patients undergoing regimen B (ifosfamide, carboplatin, and etoposide [ICE] chemotherapy) were given granulocyte colony-stimulating factor because of severe myelosuppression. The mean follow-up period was 24 months (range, 4 – 62 months) after the diagnosis of metastatic retinoblastoma was established. The prognoses of metastatic retinoblastoma patients were assessed according to the following major categories of involvement.
Central Nervous System Metastasis All nine patients had CNS masses detected on MRI. Histopathologically, 5 patients (patients 1, 3– 6) had orbital optic nerve involvement and extrascleral extension, 2 had tumor at the optic nerve cut end and extrascleral extension (patients 9 and 10), 1 had retrolaminar optic nerve involvement (patient 7), and 1 had extrascleral extension only without optic nerve involvement (patient 14). Three of 9 patients (patients 3, 7, and 10) were treated with regimen A chemotherapy, and 6 (patients 1, 4 – 6, 9, and 14) were
1560
treated with regimen B chemotherapy. All patients underwent craniospinal EBRT. At the end of follow-up, all patients had died of CNS retinoblastoma.
Concurrent Central Nervous System and Distant Metastasis In addition to CNS metastasis, 2 patients had bone marrow involvement, 2 had bone involvement, and 1 had both bone marrow and bone involvement. Histopathologically, 3 patients had retrolaminar optic nerve involvement and extrascleral extension (patients 8, 11, and 16), 1 had tumor at the optic nerve cut end and extrascleral extension (patient 2), and 1 had extrascleral extension only without optic nerve involvement (patient 13). One of 5 patients (patient 8) was treated with regimen A chemotherapy, and 4 (patients 2, 11, 13, and 16) were treated with regimen B chemotherapy. All patients underwent craniospinal EBRT. At the end of follow-up, all patients had died of diseaserelated complications.
Distant Metastasis Two patients had metastasis to the bone, 1 to the bone marrow and 1 to the lymph nodes (Fig 4). Histopathologically, 2 patients had tumor at the cut end of the optic nerve and extrascleral extension (patients 15 and 18), 1 had retrolaminar optic nerve involvement
Gündüz et al 䡠 Metastatic Retinoblastoma
Figure 1. A, Photograph showing the facial appearance of the 6-year-old boy with massive extraocular spread of retinoblastoma resulting in an hemorrhagic orbital mass. B, T1-weighted orbital magnetic resonance imaging scan showing the large retinoblastoma mass in the orbit that had destroyed the globe.
and extrascleral extension (patient 17), and 1 had extrascleral extension only (patient 12). All 4 patients were treated with regimen B chemotherapy and were alive at follow-ups of 9, 20, 37, and 62 months after treatment.
in the previous reports. There was no gender predilection for extraocular spread of retinoblastoma in our series. Only 1 patient had a family history of retinoblastoma. In our series, extraocular spread always occurred unilaterally, even in the patients with bilateral disease.
Discussion
Patient Characteristics
Patient Demographics, Laterality, and Family History
Although 5 of 7 Reese-Ellsworth group 5 patients in group 2 in our series had large tumors (with largest tumor base diameter ⬎15 mm), multivariate analysis in previous reports did not find tumor size to be a risk factor for metastasis.13,14 Iris neovascularization and neovascular glaucoma were present in 2 of 9 eyes, and pseudohypopyon was observed in 1 of 9 eyes. With regard to iris neovascularization and neovascular glaucoma, one study found that these were risk factors for extraocular spread of retinoblastoma,15 whereas another study did not reveal a similar relationship.14 Neovascular glaucoma probably causes thinning of the sclera and facilitates spread of retinoblastoma cells
The mean age of diagnosis of retinoblastoma is approximately 12 months in children who have bilateral disease and 24 months in children who have unilateral disease.12 It has been reported that the risk of developing metastatic disease is higher in patients whose clinical diagnosis of retinoblastoma is delayed.13,14 In 2 series of patients with extraocular retinoblastoma, the mean age at presentation was 33 and 38 months.6,7 In our study, the mean age at presentation was 45 months, reflecting that the diagnosis was more delayed than
1561
Ophthalmology Volume 113, Number 9, September 2006
Figure 2. A, Photograph showing the facial appearance of 7-year-old girl with an enucleated right eye and a buphthalmic hemorrhagic globe on the left side. B, T1-weighted orbital magnetic resonance imaging scan showing the thickening and infiltration of the optic nerve on the left side. C, Photograph showing the facial appearance of the child 3 months after initiation of chemotherapy. No surgical treatment was administered to the left eye.
1562
Gündüz et al 䡠 Metastatic Retinoblastoma
Figure 3. A, Photograph of a 4.5-year old boy with a massive orbital retinoblastoma. The child underwent evisceration and polymethyl methacrylate ball implantation elsewhere for nonresolving vitreous hemorrhage. B, T1-weighted orbital magnetic resonance imaging scan depicting the massive orbital retinoblastoma and the ball implant seen as a dark signal object in the orbit.
outside the eye. Pseudohypopyon was found to be a risk factor for metastasis in one study.16 Retrolaminar optic nerve invasion is reported to be uncommon in patients with diffuse infiltrating retinoblastoma17,18; however, the 2 patients (patients 8 and 11) with diffuse infiltrating retinoblastoma had retrolaminar optic nerve involvement in our series and eventual CNS involvement. In 2 patients in our series (patients 2 and 15), tumor recurrence occurred 46 months and 10 years, respectively, after EBRT for intraocular disease. Both of these patients had extrascleral spread and orbital involvement at presentation. Although retinoblastoma recurrence after EBRT generally is detected within 2 years after treatment, recurrences have been reported to occur much later, up to 25 years after
radiotherapy.19 –22 These results attest to the importance of extended follow-up in retinoblastoma patients who have been treated with primary EBRT. The 3 patients in our Group 3 were mismanaged elsewhere because retinoblastoma was not anticipated. Shields et al23 found that adjuvant chemotherapy is useful to prevent metastasis in patients with unsuspected retinoblastoma who undergo pars plana vitrectomy.
Risk Factors for Metastatic Retinoblastoma and Adjuvant Treatment Optic nerve involvement is the most important risk factor for CNS metastasis in retinoblastoma.3,24 –26 When tumor is
1563
Ophthalmology Volume 113, Number 9, September 2006 Table 3. Treatment Methods, Final Outcomes, and Duration of Follow-up in Metastatic Retinoblastoma Patients
Patient No. Group 1 1 2 3 4 5 6 Group 2 7 8 9 10 11 12 13 14 15 Group 3 16 17 18
Extraocular Extension
External Beam Radiotherapy
Metastatic Site
Chemotherapy Regimen*
Orbital
Craniospinal
Final Outcome
Follow-up (mos)
ON, ES ON, ES ON, ES ON, ES ON, ES ON, ES
CNS CNS⫹D CNS CNS CNS CNS
B B A B B B
⫹ ⫹ ⫹ ⫹ ⫹ ⫹
⫹ ⫹ ⫹ ⫹ ⫹ ⫹
DOD DOD DOD DOD DOD DOD
15 16† 25 8 4 12
RLON RLON, ES CEON, ES CEON, ES RLON, ES ES ES ES CEON, ES
CNS CNS⫹D CNS CNS CNS⫹D D CNS⫹D CNS D
A A B A B B B B B
– ⫹ ⫹ ⫹ ⫹ – – – ⫹
⫹ ⫹ ⫹ ⫹ ⫹ – ⫹ ⫹ –
DOD DOD DOD DOD DOD Alive DOD DOD Alive
31 54 11 18 13 62 40 34 37†
RLON, ES RLON, ES CEON, ES
CNS⫹D D D
B B B
– – ⫹
⫹ – –
DOD Alive Alive
23 20 9
CEON ⫽ cut-end optic nerve; CNS ⫽ central nervous system; D ⫽ distant metastasis; DOD ⫽ dead of disease; EBRT ⫽ external beam radiotherapy; ES ⫽ extrascleral extension; ON ⫽ orbital optic nerve involvement on histopathologic examination or magnetic resonance imaging scan; RLON ⫽ retrolaminar optic nerve. ⫹, given; ⫺, not given. *See Table 1. † These patients had retinoblastoma recurrence after primary EBRT.
present at the cut end of the optic nerve, adjuvant chemotherapy together with orbital radiotherapy (45 Gy) reduce the risk of metastasis.27–29 Retrolaminar optic nerve involvement also was found to be a risk factor for
CNS metastasis in many studies.3,24,27,30 In our study, 4 patients (patients 7, 8, 11, and 16) with retrolaminar optic nerve involvement experienced CNS metastasis. Adjuvant chemotherapy without orbital radiotherapy also is
Figure 4. Photograph of a 2-year-old boy with retinoblastoma metastatic to preauricular lymph node after undergoing anterior chamber tap for nonresolving hyphema. Because of progressive proptosis, the referring surgeon elected to close the eyelids via sutures tied over 2 pieces of sponge.
1564
Gündüz et al 䡠 Metastatic Retinoblastoma recommended when retrolaminar optic nerve involvement is present.27 Although optic nerve involvement is a major risk factor for CNS metastasis, CNS involvement also can develop hematogenously; therefore, it can be seen in patients who have extraocular spread without optic nerve involvement.3 In 1 patient in our series (patient 14), CNS involvement developed after extrascleral spread in the absence of any optic nerve involvement. Extrascleral extension has been reported to be the most significant risk factor for distant metastasis, as the tumor gains access to the vascular and lymphatic channels outside the eye.3 Drug combinations similar to those discussed above can be used to decrease the risk of metastasis. Massive choroidal involvement does not, by itself, seem to be a risk factor for metastasis; however, when it is associated with retrolaminar optic nerve involvement, the risk of metastasis is clearly present.24 –26,30 Treatment for orbital retinoblastoma, in which the tumor cells disseminate extrasclerally, usually includes chemotherapy and orbital radiotherapy to decrease the risk of metastasis.29,31 Exenteration should be reserved only for those patients with massive orbital involvement. Antonelli et al6 found survival rates of 55% and 59% in 2 different chemotherapy groups for extraocular retinoblastoma. Goble et al32 used a combination treatment including excisional biopsy, orbital radiotherapy, and chemotherapy in 5 patients with orbital retinoblastoma, and all 5 survived from 8 to 84 months. Other risk factors that warrant adjuvant chemotherapy include anterior chamber seeding, iris infiltration, ciliary body infiltration, and scleral infiltration.27 Honavar et al27 found that adjuvant chemotherapy reduces the risk of metastatic retinoblastoma in patients with these risk factors.
Metastatic Retinoblastoma and Its Treatment Metastatic disease can develop without or despite adjuvant treatment. In many studies of patients with CNS involvement, systemic chemotherapy and craniospinal irradiation have been used in treatment. Doz et al29 reported that 2 of 4 patients with CNS involvement survived at 19 and 51 months of follow-up after carboplatin, etoposide, and cyclophosphamide chemotherapy and cranial irradiation. However, in other reports, the survival rate was reported to be closer to 0%.3,6,7,10,28 In our series, the prognosis of patients with CNS metastasis was dismal; 9 patients with CNS metastasis and 5 patients with concurrent CNS and distant metastasis died despite chemotherapy with or without intrathecal treatment and craniospinal irradiation. Some studies have suggested that methotrexate and cytarabine delivered intrathecally do not provide any benefit.33,34 Therefore, chemotherapy protocols without intrathecal treatment have come into use. Regimen B (ICE: ifosfamide, carboplatin, etoposide), which was used in our study, includes agents with high penetration into potential retinoblastoma metastatic sites. Carboplatin has high penetration into CNS and bone marrow, and ifosfamide has high penetration into CNS. Ifosfamide has been used in a number of pediatric solid tumors, including osteosarcoma, Ewing’s
sarcoma, neuroblastoma, Wilms’ tumor, rhabdomyosarcoma, other soft tissue sarcomas, B-cell lymphoma, and retinoblastoma.35 The combination of ifosfamide with carboplatin and etoposide (ICE regimen) also has been found to be effective in a number of pediatric solid tumors, including lymphoma, neuroblastoma, soft tissue sarcomas, and Wilms’ tumor.11,36 The drugs in the ICE regimen have been used in the treatment of intraocular and extraocular retinoblastoma previously, either singly or in various combinations.4 Severe hematologic toxicity was found to be the most significant side effect of ICE chemotherapy.11,36 Previous studies found that the survival of patients with distant metastasis outside the CNS is approximately 50% at 5 years.5–7,37 Successful results were reported with consolidation high-dose intensive carboplatin, etoposide, and cyclophosphamide chemotherapy and autologous stem cell rescue, with a 3-year survival rate of 67% in one series5 and 50% in other series.37 In our study, 4 patients with distant metastasis were alive at follow-up periods ranging from 9 to 62 months after ICE chemotherapy without stem cell transplantation. This confirms the findings of previous studies that the prognosis of retinoblastoma patients with distant metastasis excluding the CNS is better than that of patients with CNS metastasis.6,7,10 All patients with metastatic retinoblastoma in our series had unilateral extraocular disease. In our experience, metastatic retinoblastoma patients with distant organ involvement seem to have a better prognosis than those having CNS or concurrent CNS and distant metastasis.
References 1. Magramm I, Abramson DH, Ellsworth RM. Optic nerve involvement in retinoblastoma. Ophthalmology 1989;96: 217–22. 2. Rootman J, Ellsworth RM, Hofbauer J, Kitchen D. Orbital extension of retinoblastoma: a clinicopathological study. Can J Ophthalmol 1978;13:72– 80. 3. Finger PT, Harbour JW, Karcioglu ZA. Risk factors for metastasis in retinoblastoma. Surv Ophthalmol 2002;47:1–16. 4. Finger PT, Czechonska G, Demirci H, Rausen A. Chemotherapy for retinoblastoma: a current topic. Drugs 1999;58:983–96. 5. Namouni F, Doz F, Tanguy ML, et al. High-dose chemotherapy with carboplatin, etoposide, and cyclophosphamide followed by haematopoietic stem cell rescue in patients with high-risk retinoblastoma: a SFOP and SFGM study. Eur J Cancer 1997;33:2368 –75. 6. Antoneli CB, Steinhorst F, de Cassia Braga Ribeiro K, et al. Extraocular retinoblastoma: a 13-year experience. Cancer 2003;98:1292– 8. 7. Chantada G, Fandiño A, Casak S, et al. Treatment of overt extraocular retinoblastoma. Med Pediatr Oncol 2003;40: 158 – 61. 8. Günalp I˙, Gündüz K, Arslan Y. Retinoblastoma in Turkey: diagnosis and clinical characteristics. Ophthalmic Genet 1996; 17:21–7. 9. Leal-Leal C, Flores-Rojo M, Medina-Sanson A, et al. A multicentre report from the Mexican Retinoblastoma Group. Br J Ophthalmol 2004;88:1074 –7. 10. Schvartzman E, Chantada G, Fandiño A, et al. Results of a stage-based protocol for the treatment of retinoblastoma. J Clin Oncol 1996;14:1532– 6.
1565
Ophthalmology Volume 113, Number 9, September 2006 11. Abu-Ghosh AM, Krailo MD, Goldman SC, et al. Ifosfamide, carboplatin and etoposide in children with poor-risk relapsed Wilms’ tumor: a Children’s Cancer Group report. Ann Oncol 2002;13:460 –9. 12. Sanders BM, Draper GJ, Kingston JE. Retinoblastoma in Great Britain 1969-80: incidence, treatment and survival. Br J Ophthalmol 1988;72:576 – 83. 13. Kopelman JE, McLean IW, Rosenberg SH. Multivariate analysis of risk factors for metastasis in retinoblastoma treated by enucleation. Ophthalmology 1987;94:371–7. 14. Baez KA, Ulbig MW, Cater J, et al. Iris neovascularization, increased intraocular pressure and vitreous hemorrhage as risk factors for invasion of the optic nerve and choroid in children with retinoblastoma [in German]. Ophthalmologe 1994;91: 796 – 800. 15. Messmer EP, Heinrich T, Hopping W, et al. Risk factors for metastases in patients with retinoblastoma. Ophthalmology 1991;98:136 – 41. 16. Wang AG, Hsu WM, Hsia WW, et al. Clinicopathologic factors related to metastasis in retinoblastoma. J Pediatric Ophthalmol Strabismus 2001;38:166 –71. 17. Materin MA, Shields CL, Shields JA, Eagle RC Jr. Diffuse infiltrating retinoblastoma simulating uveitis in a 7-year-old boy. Arch Ophthalmol 2000;118:442–3. 18. Bhatnagar R, Vine AK. Diffuse infiltrating retinoblastoma. Ophthalmology 1991;98:1657– 61. 19. Abramson DH, Servodido CA, De Lillo AR, et al. Recurrence of unilateral retinoblastoma following radiation therapy. Ophthalmic Genet 1994;15:107–13. 20. Shields CL, Piccone MR, Shields JA, et al. Mushroom-shaped choroidal recurrence of retinoblastoma 25 years after therapy. Arch Ophthalmol 2002;120:844 – 6. 21. Goto H, Kousaka A, Takano S, Usui M. Recurrence of retinoblastoma 12 years after brachytherapy. Am J Ophthalmol 2002;134:773–5. 22. Messmer EP, Sauerwein W, Heinrich T, et al. New and recurrent tumor foci following local treatment as well as external beam radiation in eyes of patients with hereditary retinoblastoma. Graefes Arch Clin Exp Ophthalmol 1990;228:426 –31. 23. Shields CL, Honavar SG, Shields JA, et al. Vitrectomy in eyes with unsuspected retinoblastoma. Ophthalmology 2000;107: 2250 –5.
1566
24. Khelfaoui F, Validire P, Auperin A, et al. Histopathologic risk factors in retinoblastoma: a retrospective study of 172 patients treated in a single institution. Cancer 1996;77:1206 –13. 25. Shields CL, Shields JA, Baez KA, et al. Choroidal invasion of retinoblastoma: metastatic potential and clinical risk factors. Br J Ophthalmol 1993;77:544 – 8. 26. Shields CL, Shields JA, Baez K, et al. Optic nerve invasion of retinoblastoma: metastatic potential and clinical risk factors. Cancer 1994;73:692– 8. 27. Honavar SG, Singh AD, Shields CL, et al. Postenucleation adjuvant therapy in high-risk retinoblastoma. Arch Ophthalmol 2002;120:923–31. 28. Zelter M, Damel A, Gonzales G, Schwartz L. A prospective study on the treatment of retinoblastoma in 72 patients. Cancer 1991;68:1685–90. 29. Doz F, Neuenschwander S, Plantaz D, et al. Etoposide and carboplatin in extraocular retinoblastoma: a study by the Societe Francaise d’Oncologie Pediatrique. J Clin Oncol 1995; 13:902–9. 30. Hungerford J. Factors influencing metastasis in retinoblastoma. Br J Ophthalmol 1993;77:541. 31. Hungerford J, Kingston J, Plowman N. Orbital recurrence of retinoblastoma. Ophthalmic Paediatr Genet 1987;8:63– 8. 32. Goble RR, McKenzie J, Kingston JE, et al. Orbital recurrence of retinoblastoma successfully treated by combined therapy. Br J Ophthalmol 1990;74:97– 8. 33. Pratt CB, Fontanesi J, Chenaille P, et al. Chemotherapy for extraocular retinoblastoma. Pediatr Hematol Oncol 1994;11: 301–9. 34. Pratt CB. Use of chemotherapy for retinoblastoma. Med Pediatr Oncol 1998;31:531–3. 35. Pratt CB, Douglass EC, Etcubanas EL, et al. Ifosfamide in pediatric malignant solid tumors. Cancer Chemother Pharmacol 1989;24(suppl):S24 –7. 36. Kung FH, Desai SJ, Dickerman JD, et al. Ifosfamide/carboplatin/etoposide (ICE) for recurrent malignant solid tumors of childhood: a Pediatric Oncology Group Phase I/II study. J Pediatr Hematol Oncol 1995;17:265–9. 37. Rodriguez-Galindo C, Wilson MW, Haik BG, et al. Treatment of metastatic retinoblastoma. Ophthalmology 2003;110:1237– 40.
Retinoblastoma is the most common intraocular malignant tumor in children. In neglected or untreated cases, retinoblastoma can demonstrate extraocular spread, primarily through the optic nerve,1 but also through the sclera.2 Orbital invasion via either route is a risk factor for metastasis.3 Retinoblastoma can metastasize to the regional lymph nodes, central nervous system (CNS), and distant organs, usually bone and bone marrow.3,4 In contrast to patients with intraocular retinoblastoma, patients with metastatic retinoblastoma have a poor prognosis.3–7 Metastatic retinoblastoma is rare in advanced countries. However, it is a major cause of retinoblastoma-related mortality in developing countries. In 2 reports from developing countries, the frequency of metastatic retinoblastoma was quoted as approximately 9%8 and 11%.9 Originally received: September 22, 2005. Accepted: March 27, 2006. Manuscript no. 2005-853. 1 Department of Ophthalmology, Ankara University Faculty of Medicine, Ankara, Turkey. 2 Department of Pediatric Oncology, Ankara University Faculty of Medicine, Ankara, Turkey. Correspondence to Kaan Gündüz, MD, Mesa Koza Plaza 24/18, Küpe Sok, GOP 06700, Ankara, Turkey. E-mail: [email protected].
1558
© 2006 by the American Academy of Ophthalmology Published by Elsevier Inc.
The purpose of our study was to evaluate the clinical features, treatment, and prognosis of patients with metastatic retinoblastoma diagnosed and treated at the Ocular and Pediatric Oncology Services, Ankara University, Ankara, Turkey.
Patients and Methods The study was a retrospective, interventional case series of 18 consecutive patients with metastatic retinoblastoma who were diagnosed and treated at our oncology services between January 1999 and January 2005. Institutional review board approval was obtained for the various surgical procedures and other treatments given. The patients’ families were informed about the expected natural course of the disease, the surgical procedures, and other treatments involved, and written consent was obtained from the parents of each patient. Data recorded included age at diagnosis, gender, hereditary pattern (familial, sporadic), and laterality (unilateral, bilateral). Each patient underwent complete ocular examination under anesthesia. Intraocular pressure and anterior segment findings, including iris neovascularization, pseudohypopyon, and funduscopic features, were noted. Intraocular tumor dimensions and Reese-Ellsworth groups of retinoblastoma were noted for each eye, if applicable (i.e., if the tumor was intraocular). Patients with intraocular tumors also underwent A- and B-mode ultrasonograISSN 0161-6420/06/$–see front matter doi:10.1016/j.ophtha.2006.03.039
Gündüz et al 䡠 Metastatic Retinoblastoma phy. All patients underwent orbital and cranial magnetic resonance imaging (MRI). Each patient had histopathologic confirmation of retinoblastoma established via exenteration, enucleation of either the currently affected or the previously affected fellow eye, or orbital biopsy. Patients with massive orbital involvement had orbital exenteration. Patients with advanced intraocular disease were treated with enucleation. In some cases, extraocular spread of retinoblastoma resulted in an orbital mass. These patients underwent an orbital biopsy procedure. Based on the histopathologic and MRI data, the status of extraocular spread (retrolaminar optic nerve involvement, tumor at cut-end of optic nerve, orbital optic nerve involvement, extrascleral involvement, or a combination thereof) was assessed. As soon as extraocular spread was determined, all patients underwent lumbar puncture, bone marrow aspiration, and bone scintigraphy for metastatic evaluation. After the diagnosis of metastatic retinoblastoma was made, systemic treatment was instituted. Systemic treatment consisted of chemotherapy (regimen A or B)10,11 and craniospinal external-beam radiotherapy (24 Gy to the cranium and 18 Gy to the spine), if necessary (Table 1). The first 4 patients enrolled in this study had regimen A chemotherapy and the following 14 had regimen B chemotherapy. Patients with orbital disease (retinoblastoma infiltrating the cut end of the optic nerve, orbital optic nerve, massive extrascleral spread, or a combination thereof) received orbital external-beam radiotherapy (EBRT) at a dose of 45 Gy. Some changes were made in the original chemotherapy regimens. In regimen A, the dose of cyclophosphamide was reduced to 20 mg/kg and cisplatin was replaced with carboplatin, the latter having less nephrotoxicity and ototoxicity.10 Patients were followed up at regular chemotherapy intervals and at 2- to 3-month intervals after the cessation of chemotherapy.
Table 1. Chemotherapy Regimens Used in Metastatic Retinoblastoma Regimen A Weeks 0, 6, 12, 18, 24, 27, 30, 33 Cyclophosphamide (20 mg/kg, day 1) Doxorubicin (0.67 mg/kg, days 1–3) Vincristine (0.05 mg/kg, day 1) Weeks 3, 9, 15, 21 Carboplatin (560 mg/m2, day 1) Etoposide (3.3 mg/kg, days 1–3) Weeks 36, 39, 42, 45, 48, 51, 54, 57 Cyclophosphamide (30 mg/kg, day 1) Vincristine (0.05 mg/kg, day 1) Intrathecal chemotherapy* (weeks 0, 1, 2, 3, 4, 5, 6, and at every 3-week chemotherapy cycle after week 6) Age ⱕ2 years: methotrexate (6.0 mg), cytosine arabinoside (12.0 mg), dexamethasone (2.0 mg/m2) Age ⬎2–3 years: methotrexate (9.0 mg), cytosine arabinoside (18.0 mg), dexamethasone (2.0 mg/m2) Age ⬎3–9 years: methotrexate (12.0 mg), cytosine arabinoside (24.0 mg), dexamethasone (2.0 mg/m2) Age ⬎9 years: methotrexate (12.0 mg), cytosine arabinoside (30.0 mg), dexamethasone (2.0 mg/m2) Regimen B (ICE ⫽ ifosfamide, carboplatin, etoposide) Weeks 1, 4, 7, 10, 13, 16 Ifosfamide (1.8 g/m2/day, days 1–5) Carboplatin (560 mg/m2/day, day 1) Etoposide (100 mg/m2/day, days 1–5) (No intrathecal chemotherapy) *In patients with central nervous system involvement.
Results Eighteen consecutive patients with metastatic retinoblastoma were included in this study. Patient demographics, ocular status, and surgical methods used are shown in Table 2. Mean patient age was 45 months (range, 13– 86 months). Ten patients (55%) were males and 8 (45%) were females. The disease was sporadic in 17 patients (94%), and 1 patient (6%) had a family history of retinoblastoma. Retinoblastoma was bilateral in 7 (38%) patients, unilateral in 10 (56%) patients, and trilateral in 1 (6%) patient. Patients with metastatic retinoblastoma were divided into 3 groups according to the ocular presentation patterns described below. Group 1 included 6 eyes of 6 patients (patients 1– 6) with clinically manifest extraocular retinoblastoma at presentation. The diagnosis of extraocular retinoblastoma was made clinically and by neuroimaging studies. The mean age at presentation was 62 months (range, 26 – 86 months). The delay in presentation of these patients was the result of physician misdiagnosis, lack of health insurance, or both, which prevented the families from seeking proper medical care. One eye had intraocular recurrence of retinoblastoma 46 months after EBRT, and the tumor subsequently spread extraocularly (patient 2). Three patients in group 1 underwent lid-sparing exenteration because of massive orbital tumor (Fig 1), and 2 patients underwent orbital biopsy to establish histopathologic diagnosis. One patient (patient 4) underwent enucleation of the fellow eye 2 years previously, and the diagnosis of retinoblastoma was established histopathologically (Fig 2A). Surgery was not carried out in the other eye of this patient. Group 2 included 9 eyes of 9 patients (patients 7–15) who had advanced intraocular disease on clinical examination and had histopathologic evidence of extraocular spread on enucleation. The mean age at presentation was 36 months (range,13–70 months). Two of 9 eyes (patients 8 and 11) had diffuse infiltrating retinoblastoma and 7 of 9 eyes (patients 7, 9, 10, 13, and 14) had Reese-Ellsworth group 5B retinoblastoma in the affected eyes. The optic disc was not visible in the affected eyes of 5 of these 7 patients, because the ocular media was hazy or the tumor was overhanging the optic nerve. The remaining 2 of 7 eyes (patients 12 and 15) had massive recurrent intraocular retinoblastoma. One patient (patient 15) had been treated with primary EBRT 10 years previously at our oncology service and had been followed up for 37 months. The patient then was lost to follow-up until the tumor recurred. The other patient (patient 12) had received chemotherapy (carboplatin) elsewhere. All 9 eyes in group 2 underwent enucleation. Group 3 was composed of 3 eyes of 3 patients who had undergone surgical intervention (evisceration, patient 16; pars plana vitrectomy, patient 17; anterior chamber washout, patient 18) elsewhere because of misdiagnosis of an unsuspected retinoblastoma (Fig 3). Patient 17 underwent pars plana vitrectomy for a retinal lesion diagnosed to be a hamartoma. The patient developed extrascleral extension at the vitrectomy port and experienced subsequent orbital involvement. The mean age at presentation to our oncology service was 39 months (range, 26 –56 months). All 3 eyes were enucleated at our service. Table 3 shows the extraocular spread status, metastatic sites, radiotherapy and chemotherapy administrations, follow-up period, and final patient status. Based on the histopathologic and MRI data, all 18 patients with metastatic retinoblastoma had extraocular extension. Five of 18 patients (patients 7, 8, 11, 16, and 17) had retrolaminar optic nerve involvement, 4 (patients 9, 10, 15, and 18) had tumor at the optic nerve cut end, 6 (patients 1– 6) had tumor in the orbital optic nerve, and 17 (all patients except patient 17) had extrascleral extension. Extraocular extension occurred unilaterally in every case. Twelve of 18 patients (patients 1– 6, 8 –11, 15, 18) had massive orbital retinoblastoma and received orbital radiotherapy (45 Gy).
1559
Ophthalmology Volume 113, Number 9, September 2006 Table 2. Demographics and Examination Results of Metastatic Retinoblastoma Patients at Presentation Age at Patient Diagnosis No. Gender (yrs) Laterality
Ocular Status
Surgery
Group 1 (manifest extraocular orbital disease) 1 M 4.2 B Buphthalmos, orbital mass, massive proptosis 2 F 4.8 B Enlarged and perforated globe, orbital mass, massive proptosis (recurrence 46 mos after EBRT) 3 M 5.8 U Buphthalmos, hemorrhagic orbital mass 4 F 7.2 B Buphthalmos, orbital mass 5 F 2.2 T Buphthalmos, orbital mass 6 M 7.0 U Buphthalmos, orbital mass diagnosed as endophthalmitis elsewhere
EXE EXE EXE E (fellow eye) Biopsy Biopsy
Anterior Segment Findings Tumor Dimension (mm) Group 2 (advanced intraocular disease) 7 F 5.8 U
Pseudohypyon
Iris Neovascularization
Neovascular Glaucoma
Reese–Ellsworth Classification
–
–
–
5B
E
– – – ⫹ – – – –
– ⫹ ⫹ ⫹ – – – –
– – ⫹ ⫹
N/A 5B 5B N/A 5B 5B 5B 5B
E E E E E E E E
5.7 2.5 2.2 2.3 1.6 1.1 2.8 2.7
U B U U B U U B
Tumor filling vitreous cavity DIR 30 ⫻ 25 ⫻ 12 30 ⫻ 20 ⫻ 7 DIR 7.5 ⫻ 6.5 ⫻ 3 15 ⫻ 10 ⫻ 8 6⫻6⫻2 25 ⫻ 15 ⫻ 8
Group 3 (misdiagnosis) 16 M 4.7 17 M 2.9 18 M 2.2
U U B
S/P Evisceration due to non-resolving VH (elsewhere) S/P PPV for presumed hamartoma (elsewhere) S/P Anterior chamber tap for non-resolving hyphema (elsewhere)
8 9 10 11 12 13 14 15
M F F M F F M M
E E E
B ⫽ bilateral; DIR ⫽ diffuse infiltrating retinoblastoma; E ⫽ enucleation; EBRT ⫽ external beam radiotherapy; EXE ⫽ exenteration; F ⫽ female; M ⫽ male; N/A ⫽ not available; PPV ⫽ pars plana vitrectomy; Pseudohpy. ⫽ pseudohypopyon; S/P ⫽ status post; T ⫽ trilateral; U ⫽ unilateral; VH ⫽ vitreous hemorrhage. ⫹, present; ⫺, absent.
Systemic evaluation showed that 9 of 18 patients (patients 1, 3– 6, 7, 9, 10, and 14) had CNS metastasis, 5 (patients 2, 8, 11, 13, and 16) had concurrent CNS and distant organ metastasis, and 4 (patients 12, 15, 17, and 18) had distant organ metastasis. All patients underwent systemic chemotherapy (regimen A or B; Fig 2). The major complication of systemic chemotherapy was myelosuppression and occasional episodes of neutropenic fever, which occurred in all patients and was treated appropriately. Patients undergoing regimen B (ifosfamide, carboplatin, and etoposide [ICE] chemotherapy) were given granulocyte colony-stimulating factor because of severe myelosuppression. The mean follow-up period was 24 months (range, 4 – 62 months) after the diagnosis of metastatic retinoblastoma was established. The prognoses of metastatic retinoblastoma patients were assessed according to the following major categories of involvement.
Central Nervous System Metastasis All nine patients had CNS masses detected on MRI. Histopathologically, 5 patients (patients 1, 3– 6) had orbital optic nerve involvement and extrascleral extension, 2 had tumor at the optic nerve cut end and extrascleral extension (patients 9 and 10), 1 had retrolaminar optic nerve involvement (patient 7), and 1 had extrascleral extension only without optic nerve involvement (patient 14). Three of 9 patients (patients 3, 7, and 10) were treated with regimen A chemotherapy, and 6 (patients 1, 4 – 6, 9, and 14) were
1560
treated with regimen B chemotherapy. All patients underwent craniospinal EBRT. At the end of follow-up, all patients had died of CNS retinoblastoma.
Concurrent Central Nervous System and Distant Metastasis In addition to CNS metastasis, 2 patients had bone marrow involvement, 2 had bone involvement, and 1 had both bone marrow and bone involvement. Histopathologically, 3 patients had retrolaminar optic nerve involvement and extrascleral extension (patients 8, 11, and 16), 1 had tumor at the optic nerve cut end and extrascleral extension (patient 2), and 1 had extrascleral extension only without optic nerve involvement (patient 13). One of 5 patients (patient 8) was treated with regimen A chemotherapy, and 4 (patients 2, 11, 13, and 16) were treated with regimen B chemotherapy. All patients underwent craniospinal EBRT. At the end of follow-up, all patients had died of diseaserelated complications.
Distant Metastasis Two patients had metastasis to the bone, 1 to the bone marrow and 1 to the lymph nodes (Fig 4). Histopathologically, 2 patients had tumor at the cut end of the optic nerve and extrascleral extension (patients 15 and 18), 1 had retrolaminar optic nerve involvement
Gündüz et al 䡠 Metastatic Retinoblastoma
Figure 1. A, Photograph showing the facial appearance of the 6-year-old boy with massive extraocular spread of retinoblastoma resulting in an hemorrhagic orbital mass. B, T1-weighted orbital magnetic resonance imaging scan showing the large retinoblastoma mass in the orbit that had destroyed the globe.
and extrascleral extension (patient 17), and 1 had extrascleral extension only (patient 12). All 4 patients were treated with regimen B chemotherapy and were alive at follow-ups of 9, 20, 37, and 62 months after treatment.
in the previous reports. There was no gender predilection for extraocular spread of retinoblastoma in our series. Only 1 patient had a family history of retinoblastoma. In our series, extraocular spread always occurred unilaterally, even in the patients with bilateral disease.
Discussion
Patient Characteristics
Patient Demographics, Laterality, and Family History
Although 5 of 7 Reese-Ellsworth group 5 patients in group 2 in our series had large tumors (with largest tumor base diameter ⬎15 mm), multivariate analysis in previous reports did not find tumor size to be a risk factor for metastasis.13,14 Iris neovascularization and neovascular glaucoma were present in 2 of 9 eyes, and pseudohypopyon was observed in 1 of 9 eyes. With regard to iris neovascularization and neovascular glaucoma, one study found that these were risk factors for extraocular spread of retinoblastoma,15 whereas another study did not reveal a similar relationship.14 Neovascular glaucoma probably causes thinning of the sclera and facilitates spread of retinoblastoma cells
The mean age of diagnosis of retinoblastoma is approximately 12 months in children who have bilateral disease and 24 months in children who have unilateral disease.12 It has been reported that the risk of developing metastatic disease is higher in patients whose clinical diagnosis of retinoblastoma is delayed.13,14 In 2 series of patients with extraocular retinoblastoma, the mean age at presentation was 33 and 38 months.6,7 In our study, the mean age at presentation was 45 months, reflecting that the diagnosis was more delayed than
1561
Ophthalmology Volume 113, Number 9, September 2006
Figure 2. A, Photograph showing the facial appearance of 7-year-old girl with an enucleated right eye and a buphthalmic hemorrhagic globe on the left side. B, T1-weighted orbital magnetic resonance imaging scan showing the thickening and infiltration of the optic nerve on the left side. C, Photograph showing the facial appearance of the child 3 months after initiation of chemotherapy. No surgical treatment was administered to the left eye.
1562
Gündüz et al 䡠 Metastatic Retinoblastoma
Figure 3. A, Photograph of a 4.5-year old boy with a massive orbital retinoblastoma. The child underwent evisceration and polymethyl methacrylate ball implantation elsewhere for nonresolving vitreous hemorrhage. B, T1-weighted orbital magnetic resonance imaging scan depicting the massive orbital retinoblastoma and the ball implant seen as a dark signal object in the orbit.
outside the eye. Pseudohypopyon was found to be a risk factor for metastasis in one study.16 Retrolaminar optic nerve invasion is reported to be uncommon in patients with diffuse infiltrating retinoblastoma17,18; however, the 2 patients (patients 8 and 11) with diffuse infiltrating retinoblastoma had retrolaminar optic nerve involvement in our series and eventual CNS involvement. In 2 patients in our series (patients 2 and 15), tumor recurrence occurred 46 months and 10 years, respectively, after EBRT for intraocular disease. Both of these patients had extrascleral spread and orbital involvement at presentation. Although retinoblastoma recurrence after EBRT generally is detected within 2 years after treatment, recurrences have been reported to occur much later, up to 25 years after
radiotherapy.19 –22 These results attest to the importance of extended follow-up in retinoblastoma patients who have been treated with primary EBRT. The 3 patients in our Group 3 were mismanaged elsewhere because retinoblastoma was not anticipated. Shields et al23 found that adjuvant chemotherapy is useful to prevent metastasis in patients with unsuspected retinoblastoma who undergo pars plana vitrectomy.
Risk Factors for Metastatic Retinoblastoma and Adjuvant Treatment Optic nerve involvement is the most important risk factor for CNS metastasis in retinoblastoma.3,24 –26 When tumor is
1563
Ophthalmology Volume 113, Number 9, September 2006 Table 3. Treatment Methods, Final Outcomes, and Duration of Follow-up in Metastatic Retinoblastoma Patients
Patient No. Group 1 1 2 3 4 5 6 Group 2 7 8 9 10 11 12 13 14 15 Group 3 16 17 18
Extraocular Extension
External Beam Radiotherapy
Metastatic Site
Chemotherapy Regimen*
Orbital
Craniospinal
Final Outcome
Follow-up (mos)
ON, ES ON, ES ON, ES ON, ES ON, ES ON, ES
CNS CNS⫹D CNS CNS CNS CNS
B B A B B B
⫹ ⫹ ⫹ ⫹ ⫹ ⫹
⫹ ⫹ ⫹ ⫹ ⫹ ⫹
DOD DOD DOD DOD DOD DOD
15 16† 25 8 4 12
RLON RLON, ES CEON, ES CEON, ES RLON, ES ES ES ES CEON, ES
CNS CNS⫹D CNS CNS CNS⫹D D CNS⫹D CNS D
A A B A B B B B B
– ⫹ ⫹ ⫹ ⫹ – – – ⫹
⫹ ⫹ ⫹ ⫹ ⫹ – ⫹ ⫹ –
DOD DOD DOD DOD DOD Alive DOD DOD Alive
31 54 11 18 13 62 40 34 37†
RLON, ES RLON, ES CEON, ES
CNS⫹D D D
B B B
– – ⫹
⫹ – –
DOD Alive Alive
23 20 9
CEON ⫽ cut-end optic nerve; CNS ⫽ central nervous system; D ⫽ distant metastasis; DOD ⫽ dead of disease; EBRT ⫽ external beam radiotherapy; ES ⫽ extrascleral extension; ON ⫽ orbital optic nerve involvement on histopathologic examination or magnetic resonance imaging scan; RLON ⫽ retrolaminar optic nerve. ⫹, given; ⫺, not given. *See Table 1. † These patients had retinoblastoma recurrence after primary EBRT.
present at the cut end of the optic nerve, adjuvant chemotherapy together with orbital radiotherapy (45 Gy) reduce the risk of metastasis.27–29 Retrolaminar optic nerve involvement also was found to be a risk factor for
CNS metastasis in many studies.3,24,27,30 In our study, 4 patients (patients 7, 8, 11, and 16) with retrolaminar optic nerve involvement experienced CNS metastasis. Adjuvant chemotherapy without orbital radiotherapy also is
Figure 4. Photograph of a 2-year-old boy with retinoblastoma metastatic to preauricular lymph node after undergoing anterior chamber tap for nonresolving hyphema. Because of progressive proptosis, the referring surgeon elected to close the eyelids via sutures tied over 2 pieces of sponge.
1564
Gündüz et al 䡠 Metastatic Retinoblastoma recommended when retrolaminar optic nerve involvement is present.27 Although optic nerve involvement is a major risk factor for CNS metastasis, CNS involvement also can develop hematogenously; therefore, it can be seen in patients who have extraocular spread without optic nerve involvement.3 In 1 patient in our series (patient 14), CNS involvement developed after extrascleral spread in the absence of any optic nerve involvement. Extrascleral extension has been reported to be the most significant risk factor for distant metastasis, as the tumor gains access to the vascular and lymphatic channels outside the eye.3 Drug combinations similar to those discussed above can be used to decrease the risk of metastasis. Massive choroidal involvement does not, by itself, seem to be a risk factor for metastasis; however, when it is associated with retrolaminar optic nerve involvement, the risk of metastasis is clearly present.24 –26,30 Treatment for orbital retinoblastoma, in which the tumor cells disseminate extrasclerally, usually includes chemotherapy and orbital radiotherapy to decrease the risk of metastasis.29,31 Exenteration should be reserved only for those patients with massive orbital involvement. Antonelli et al6 found survival rates of 55% and 59% in 2 different chemotherapy groups for extraocular retinoblastoma. Goble et al32 used a combination treatment including excisional biopsy, orbital radiotherapy, and chemotherapy in 5 patients with orbital retinoblastoma, and all 5 survived from 8 to 84 months. Other risk factors that warrant adjuvant chemotherapy include anterior chamber seeding, iris infiltration, ciliary body infiltration, and scleral infiltration.27 Honavar et al27 found that adjuvant chemotherapy reduces the risk of metastatic retinoblastoma in patients with these risk factors.
Metastatic Retinoblastoma and Its Treatment Metastatic disease can develop without or despite adjuvant treatment. In many studies of patients with CNS involvement, systemic chemotherapy and craniospinal irradiation have been used in treatment. Doz et al29 reported that 2 of 4 patients with CNS involvement survived at 19 and 51 months of follow-up after carboplatin, etoposide, and cyclophosphamide chemotherapy and cranial irradiation. However, in other reports, the survival rate was reported to be closer to 0%.3,6,7,10,28 In our series, the prognosis of patients with CNS metastasis was dismal; 9 patients with CNS metastasis and 5 patients with concurrent CNS and distant metastasis died despite chemotherapy with or without intrathecal treatment and craniospinal irradiation. Some studies have suggested that methotrexate and cytarabine delivered intrathecally do not provide any benefit.33,34 Therefore, chemotherapy protocols without intrathecal treatment have come into use. Regimen B (ICE: ifosfamide, carboplatin, etoposide), which was used in our study, includes agents with high penetration into potential retinoblastoma metastatic sites. Carboplatin has high penetration into CNS and bone marrow, and ifosfamide has high penetration into CNS. Ifosfamide has been used in a number of pediatric solid tumors, including osteosarcoma, Ewing’s
sarcoma, neuroblastoma, Wilms’ tumor, rhabdomyosarcoma, other soft tissue sarcomas, B-cell lymphoma, and retinoblastoma.35 The combination of ifosfamide with carboplatin and etoposide (ICE regimen) also has been found to be effective in a number of pediatric solid tumors, including lymphoma, neuroblastoma, soft tissue sarcomas, and Wilms’ tumor.11,36 The drugs in the ICE regimen have been used in the treatment of intraocular and extraocular retinoblastoma previously, either singly or in various combinations.4 Severe hematologic toxicity was found to be the most significant side effect of ICE chemotherapy.11,36 Previous studies found that the survival of patients with distant metastasis outside the CNS is approximately 50% at 5 years.5–7,37 Successful results were reported with consolidation high-dose intensive carboplatin, etoposide, and cyclophosphamide chemotherapy and autologous stem cell rescue, with a 3-year survival rate of 67% in one series5 and 50% in other series.37 In our study, 4 patients with distant metastasis were alive at follow-up periods ranging from 9 to 62 months after ICE chemotherapy without stem cell transplantation. This confirms the findings of previous studies that the prognosis of retinoblastoma patients with distant metastasis excluding the CNS is better than that of patients with CNS metastasis.6,7,10 All patients with metastatic retinoblastoma in our series had unilateral extraocular disease. In our experience, metastatic retinoblastoma patients with distant organ involvement seem to have a better prognosis than those having CNS or concurrent CNS and distant metastasis.
References 1. Magramm I, Abramson DH, Ellsworth RM. Optic nerve involvement in retinoblastoma. Ophthalmology 1989;96: 217–22. 2. Rootman J, Ellsworth RM, Hofbauer J, Kitchen D. Orbital extension of retinoblastoma: a clinicopathological study. Can J Ophthalmol 1978;13:72– 80. 3. Finger PT, Harbour JW, Karcioglu ZA. Risk factors for metastasis in retinoblastoma. Surv Ophthalmol 2002;47:1–16. 4. Finger PT, Czechonska G, Demirci H, Rausen A. Chemotherapy for retinoblastoma: a current topic. Drugs 1999;58:983–96. 5. Namouni F, Doz F, Tanguy ML, et al. High-dose chemotherapy with carboplatin, etoposide, and cyclophosphamide followed by haematopoietic stem cell rescue in patients with high-risk retinoblastoma: a SFOP and SFGM study. Eur J Cancer 1997;33:2368 –75. 6. Antoneli CB, Steinhorst F, de Cassia Braga Ribeiro K, et al. Extraocular retinoblastoma: a 13-year experience. Cancer 2003;98:1292– 8. 7. Chantada G, Fandiño A, Casak S, et al. Treatment of overt extraocular retinoblastoma. Med Pediatr Oncol 2003;40: 158 – 61. 8. Günalp I˙, Gündüz K, Arslan Y. Retinoblastoma in Turkey: diagnosis and clinical characteristics. Ophthalmic Genet 1996; 17:21–7. 9. Leal-Leal C, Flores-Rojo M, Medina-Sanson A, et al. A multicentre report from the Mexican Retinoblastoma Group. Br J Ophthalmol 2004;88:1074 –7. 10. Schvartzman E, Chantada G, Fandiño A, et al. Results of a stage-based protocol for the treatment of retinoblastoma. J Clin Oncol 1996;14:1532– 6.
1565
Ophthalmology Volume 113, Number 9, September 2006 11. Abu-Ghosh AM, Krailo MD, Goldman SC, et al. Ifosfamide, carboplatin and etoposide in children with poor-risk relapsed Wilms’ tumor: a Children’s Cancer Group report. Ann Oncol 2002;13:460 –9. 12. Sanders BM, Draper GJ, Kingston JE. Retinoblastoma in Great Britain 1969-80: incidence, treatment and survival. Br J Ophthalmol 1988;72:576 – 83. 13. Kopelman JE, McLean IW, Rosenberg SH. Multivariate analysis of risk factors for metastasis in retinoblastoma treated by enucleation. Ophthalmology 1987;94:371–7. 14. Baez KA, Ulbig MW, Cater J, et al. Iris neovascularization, increased intraocular pressure and vitreous hemorrhage as risk factors for invasion of the optic nerve and choroid in children with retinoblastoma [in German]. Ophthalmologe 1994;91: 796 – 800. 15. Messmer EP, Heinrich T, Hopping W, et al. Risk factors for metastases in patients with retinoblastoma. Ophthalmology 1991;98:136 – 41. 16. Wang AG, Hsu WM, Hsia WW, et al. Clinicopathologic factors related to metastasis in retinoblastoma. J Pediatric Ophthalmol Strabismus 2001;38:166 –71. 17. Materin MA, Shields CL, Shields JA, Eagle RC Jr. Diffuse infiltrating retinoblastoma simulating uveitis in a 7-year-old boy. Arch Ophthalmol 2000;118:442–3. 18. Bhatnagar R, Vine AK. Diffuse infiltrating retinoblastoma. Ophthalmology 1991;98:1657– 61. 19. Abramson DH, Servodido CA, De Lillo AR, et al. Recurrence of unilateral retinoblastoma following radiation therapy. Ophthalmic Genet 1994;15:107–13. 20. Shields CL, Piccone MR, Shields JA, et al. Mushroom-shaped choroidal recurrence of retinoblastoma 25 years after therapy. Arch Ophthalmol 2002;120:844 – 6. 21. Goto H, Kousaka A, Takano S, Usui M. Recurrence of retinoblastoma 12 years after brachytherapy. Am J Ophthalmol 2002;134:773–5. 22. Messmer EP, Sauerwein W, Heinrich T, et al. New and recurrent tumor foci following local treatment as well as external beam radiation in eyes of patients with hereditary retinoblastoma. Graefes Arch Clin Exp Ophthalmol 1990;228:426 –31. 23. Shields CL, Honavar SG, Shields JA, et al. Vitrectomy in eyes with unsuspected retinoblastoma. Ophthalmology 2000;107: 2250 –5.
1566
24. Khelfaoui F, Validire P, Auperin A, et al. Histopathologic risk factors in retinoblastoma: a retrospective study of 172 patients treated in a single institution. Cancer 1996;77:1206 –13. 25. Shields CL, Shields JA, Baez KA, et al. Choroidal invasion of retinoblastoma: metastatic potential and clinical risk factors. Br J Ophthalmol 1993;77:544 – 8. 26. Shields CL, Shields JA, Baez K, et al. Optic nerve invasion of retinoblastoma: metastatic potential and clinical risk factors. Cancer 1994;73:692– 8. 27. Honavar SG, Singh AD, Shields CL, et al. Postenucleation adjuvant therapy in high-risk retinoblastoma. Arch Ophthalmol 2002;120:923–31. 28. Zelter M, Damel A, Gonzales G, Schwartz L. A prospective study on the treatment of retinoblastoma in 72 patients. Cancer 1991;68:1685–90. 29. Doz F, Neuenschwander S, Plantaz D, et al. Etoposide and carboplatin in extraocular retinoblastoma: a study by the Societe Francaise d’Oncologie Pediatrique. J Clin Oncol 1995; 13:902–9. 30. Hungerford J. Factors influencing metastasis in retinoblastoma. Br J Ophthalmol 1993;77:541. 31. Hungerford J, Kingston J, Plowman N. Orbital recurrence of retinoblastoma. Ophthalmic Paediatr Genet 1987;8:63– 8. 32. Goble RR, McKenzie J, Kingston JE, et al. Orbital recurrence of retinoblastoma successfully treated by combined therapy. Br J Ophthalmol 1990;74:97– 8. 33. Pratt CB, Fontanesi J, Chenaille P, et al. Chemotherapy for extraocular retinoblastoma. Pediatr Hematol Oncol 1994;11: 301–9. 34. Pratt CB. Use of chemotherapy for retinoblastoma. Med Pediatr Oncol 1998;31:531–3. 35. Pratt CB, Douglass EC, Etcubanas EL, et al. Ifosfamide in pediatric malignant solid tumors. Cancer Chemother Pharmacol 1989;24(suppl):S24 –7. 36. Kung FH, Desai SJ, Dickerman JD, et al. Ifosfamide/carboplatin/etoposide (ICE) for recurrent malignant solid tumors of childhood: a Pediatric Oncology Group Phase I/II study. J Pediatr Hematol Oncol 1995;17:265–9. 37. Rodriguez-Galindo C, Wilson MW, Haik BG, et al. Treatment of metastatic retinoblastoma. Ophthalmology 2003;110:1237– 40.