Principles & Practice of Pediatric Oncology
5th Edition

28
Retinoblastoma
Richard L. Hurwitz
Carol L. Shields
Jerry A. Shields
Patricia Chevez-Barrios
Mary Y. Hurwitz
Murali M. Chintagumpala
Retinoblastoma is a malignant tumor of the embryonic neural retina and is the most common intraocular malignancy in children. Although usually not recognized at birth, retinoblastoma is often congenital and affects predominantly young children. The tumor has a variable growth rate, can originate from single or multiple foci in one or both eyes, and, in bilateral cases, may be manifest in one eye many months before it is evident in the other. Retinoblastoma is caused by a mutation in a gene that expresses a protein central to the control of the cell cycle and may occur sporadically or be inherited. Children with the hereditary type of retinoblastoma have a particular susceptibility to developing other malignant tumors. This disease serves as a model for understanding the genetics and heredity of childhood cancer.
The term retinoblastoma was first adopted by the American Ophthalmological Society in 1926.1 The cellular origin of retinoblastoma had been a topic of debate since 1809, when the Scottish surgeon Wardrop first recognized, based only on gross pathologic findings, that retinoblastoma is a discrete tumor arising from the retina.2,3 After this publication, other pathologists, including Robin and Langenbeck, confirmed the observations at a microscopic level. Virchow, however, thought that the cell of origin was glial and named it glioma of the retina. In the late 1800s, the term neuroepithelioma was proposed by Flexner and supported later by Wintersteiner because they believed that the tumor originated from the neuroepithelium and that the typical rosettes that now bear their names were attempts to form photoreceptors. In the early 1900s, Verhoeff concluded that the tumor was derived from undifferentiated embryonic retinal cells called retinoblasts and proposed the term retinoblastoma.1 Zimmerman proposed the term retinocytoma for the well-differentiated tumor that displays benign features.4 For this same tumor Gallie et al.5 described the clinical features and proposed the term retinoma. The histopathologic, ultrastructural, immunohistochemical, and molecular characteristics of retinoblastoma support the concept that this tumor originates from a multipotent precursor cell. This cell could develop into almost any type of inner or outer retinal cell including photoreceptors.6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21
EPIDEMIOLOGY
The Third National Cancer Survey indicates an average incidence of 11 new cases of retinoblastoma per million population younger than 5 years of age or 1 in 18,000 live births in the United States.22 Although data from developing countries are less complete, oncologists in Central and South America, the Middle East, and India generally feel that the incidence may be greater in these regions. A multicenter report from Mexico concluded that retinoblastoma is the second most frequent solid malignancy in children [after central nervous system (CNS) tumors].23 The estimated frequency of bilateral retinoblastoma ranges from 20% to 30%. Thus, in the United States, an estimated 200 children per year develop retinoblastoma; of these 200, at least 40 to 60 cases are bilateral. There are no racial or gender predilections.
Retinoblastoma is often present at birth and is almost entirely restricted to early childhood. About 80% of cases are diagnosed before 3 to 4 years of age, with a median age at diagnosis of 2 years.24 The discovery of retinoblastoma beyond the age of 6 years is rare. Bilateral disease is diagnosed earlier than unilateral disease. Sporadic bilateral retinoblastoma has been associated with advanced parental age.25
Multiple congenital anomalies associated with retinoblastoma have been reported in approximately 0.05% of U.S. patients with retinoblastoma.26 The reported anomalies include congenital cardiovascular defects, cleft palate, infantile cortical hyperostosis, dentinogenesis imperfecta, familial congenital cataracts, and incontinentia pigmenti or Bloch-Sulzberger syndrome (an X-linked inherited disease that is lethal in males but affects females with pigmentary retinopathy, corneal opacities, cataracts, nystagmus, blue sclerae, myopia, pseudoglioma, dental abnormalities, abnormal skin pigmentation, and mental deficiency).24 An association with mental retardation has been suggested in children with the D-deletion syndrome; however, most patients with retinoblastoma have no intellectual impairment.
P.866

GENETICS
The majority of retinoblastomas appear sporadically; however, an inherited form of the disease has been documented27 and is transmitted with few exceptions as a typical mendelian autosomal dominant trait with high but incomplete penetrance. Of all cases, about 60% are nonhereditary and unilateral, 15% are hereditary and unilateral, and 25% are hereditary and bilateral.28,29
A “two-hit” model has been proposed to explain the observations that familial cases are generally multifocal and bilateral, whereas sporadic cases typically present with unilateral unifocal disease at a later age.28,30 According to the model, as few as two stochastic mutational events are required for tumor initiation, the first of which can be inherited through the germ line (in heritable cases) or can occur somatically in individual retinal cells (in nonheritable cases). The second event occurs somatically in either case and leads to tumor formation from each doubly defective retinal cell.
The presence of a microscopically visible deletion in one chromosome 13 homolog in constitutional cells of a small number of retinoblastoma patients was the first evidence that supported an inherited mechanism for retinoblastoma development.31,32,33,34 Although the deletions varied between families, each deletion minimally encompassed chromosome 13, band q14.35,36 This chromosomal locus contains the RB1 retinoblastoma gene. In the two-hit model, such deletions in the germ line could act as the first hit and confer the risk of tumor formation as an autosomal dominant trait. The increasing resolution of cytogenetic technology and the development of DNA probes for loci in the immediate vicinity of the RB1 gene locus has allowed the detection of more subtle genomic rearrangements. These techniques can be used to identify people who carry nonpenetrant mutations in the retinoblastoma susceptibility locus.37
Patients without a gross thirteenth chromosomal deletion but who have bilateral or familial retinoblastoma have submicroscopic mutations at the RB1 locus similar to mutations that have been found in the tumor cells of patients with nonhereditary retinoblastoma. The second step in tumorigenesis in both heritable and nonhereditary retinoblastoma involves somatic alteration of the normal allele at the RB1 locus in such a way that the mutant allele is unmasked. Thus, the first mutation in this process, although it may be inherited as an autosomal dominant trait in the child, is in fact a recessive defect in the individual retinal cell. Elimination of the chromosome containing the wild-type allele followed by reduplication of the remaining mutant chromosome may be one mechanism by which the affected RB1 locus becomes homozygous within the cell.38,39 The potential tumor cell becomes recessive for the mutant allele.
Although the unmasking of predisposing mutations at the RB1 locus occurs in mechanistically similar ways in sporadic and heritable retinoblastoma, only the latter carries the initial mutation in each cell. Patients with heritable disease also seem to be at greatly increased risk for the development of second primary tumors, particularly osteogenic sarcoma.40 This high propensity is genetically determined by the predisposing RB1 mutation. The notion of a pathogenetic association between these two rare tumor types was tested by determining the constitutional and osteosarcoma genotypes at restriction fragment length polymorphism (RFLP) loci on chromosome 13. The data indicated that osteosarcomas arising in patients with retinoblastoma had become homozygous specifically around the chromosomal region carrying the RB1 locus.41 Furthermore, these same chromosomal mechanisms eliciting losses of constitutional heterozygosity were observed in sporadic osteosarcomas, suggesting a genetic similarity in pathogenetic causality.
These studies provided data useful for the molecular isolation of the RB1 gene.42 The genomic organization of the approximately 200-kb locus was determined and the expression of its 4.7-kb messenger RNA (mRNA) transcript in tumor and normal tissues was documented. Introduction of the wild-type gene into retinoblastoma and osteosarcoma cell lines using recombinant retroviral vector transfer resulted in a partial reversal of the tumorigenic phenotype.43,44 Further characterization of the complete RB1 genomic sequence45 allowed a rigorous cataloging of the different mutations affecting the gene in retinoblastoma tumors. More than 200 disease-causing mutations have been identified in the retinoblastoma genes of patients.37,46,47,48,49,50,51,52,53,54,55
The examination of sporadic cases of bilateral retinoblastoma showed that disease frequently arises subsequent to a new germ-line mutation in the paternal allele followed by somatic alteration or loss of the maternally derived wild-type allele.56,57 This finding suggests either that mutations in the RB1 locus occur more commonly during spermatogenesis or that the paternal chromosome in the early embryo is at a higher risk of mutation. Analyses of sporadic osteosarcomas also showed preferential mutation of the paternal allele.58
Investigations of RB1 gene alterations at both the DNA and the RNA level cumulatively reveal a strong correlative relationship between the lack of RB1 gene product and the appearance of retinoblastoma tumors. In addition to osteosarcomas, other tumor types contain mutations involving the retinoblastoma gene. Molecular analyses of small cell lung carcinomas have revealed RB1 structural abnormalities in approximately 15% of cases.59 Loss of heterozygosity for chromosome 13 has been detected in about 25% of breast cancers and related breast cancer–derived cell lines.60,61 However, a more detailed analysis of the effects of chromosome 13 mutations in tumors has been compiled and clearly shows that not all tumors are either a direct or an indirect result of loss of heterozygosity of the RB1 locus.62 The cumulative data suggest that only subsets of tumors may share a common pathogenetic mechanism that results from unmasking mutations affecting the tumor-suppressing function of RB1.
RB1 mRNA is a 4.7-kb transcript in normal human and rat tissues, including brain, kidney, ovary, spleen, liver, placenta, and retina.63 The expressed protein contains 928 amino acids and has an estimated molecular mass of 110 kd. Although the number of different types of tumors that occur as a result of inherited mutations of the RB1 locus is small, the broad tissue expression and species conservation of this gene suggest a common and potentially important role in the growth or differentiation of many cell types.
The protein has been shown to be primarily localized in the cell nucleus.64 Post-translational phosphorylation of the RB protein in quiescent cells overrides growth suppression and allows cell division to take place.65 The RB protein
P.867

also has a role in the regulation of the cell cycle of actively dividing cells. The unphosphorylated RB protein (p110RB) has been shown to bind E2F1, a transcription factor and a cell cycle regulator during the G1 stage of the cell cycle. The RB/E2F1 complex masks the E2F1 transactivation domain and inhibits surrounding enhancer elements, thereby causing transcription of E2F1-regulated genes to cease. The RB protein accomplishes this by physically associating with a histone deacetylase (HDAC1). This recruitment of the deacetylase to the E2F1 regulating domain by RB allows for deacetylation of histone, thereby modulating the local structure of the chromatin.66,67 Phosphorylation of the RB protein at the G1/S boundary results in the release of these transcriptional factors, allowing them to become positive transcriptional elements. Additional cell cycle–specific kinases become activated and facilitate the progression of the cells through G2 and M. At the completion of the cell cycle, phosphatases dephosphorylate the RB protein, allowing the protein to again sequester E2F1 and form an inactive complex. Thus, positive and negative regulation of transcription and, therefore, cell proliferation are linked to the phosphorylation cycle of the RB protein. In tumors in which RB protein is mutated or absent, these intracellular transcriptional elements are dissociated and free to promote consistent and uncontrolled progression through the cell cycle. Such behavior results in unchecked cell proliferation consistent with a malignant phenotype.
The viral oncoproteins of polyomaviruses (SV40), adenoviruses (Ad-2 and Ad-5), and papillomaviruses (HPV-16) have also been shown to complex with the RB protein.68,69,70 Because one function of these viral oncoproteins appears to be the creation of a cellular environment that is permissive for DNA synthesis, one of their modes of action may involve sequestration of the antiproliferative unphosphorylated RB protein. Releasing the cell from its negative regulation by RB might allow the cell to enter S phase and synthesize DNA. Taken together, the data support a model in which the unphosphorylated form of RB is the species active in growth suppression.
Genetic Counseling
Approximately 40% of patients with retinoblastoma have the inherited form of the disease. Because the inherited form of retinoblastoma is transmitted as an autosomal dominant trait with high but incomplete penetrance, there is a 45% chance that any given child of the patient will inherit the disease. In addition, although there is high penetrance of the retinoblastoma phenotype, the possibility exists that one of the patient’s siblings could also develop retinoblastoma even if neither of the parents were affected by the disease owing to germ-line mosaicism and low-penetrant alleles (see Chapter 2).71,72,73 All children with a family history of retinoblastoma should be screened shortly after birth by a qualified ophthalmologist to permit early detection of the disease and increase the chance of ocular and vision salvage. These increased familial risks support the need for expert genetic counseling.
To effectively counsel patients and families of retinoblastoma patients, the underlying cause of the disease must be determined. Patients who present with bilateral disease can be assumed to have a germ-line mutation in the RB1 gene. Patients with unilateral disease at presentation may also have an underlying germ-line mutation. If a mutation in the RB1 gene is detected in the tumor, somatic cells should also be screened. A mutation initially detected in the somatic cells is presumptive evidence of a mutation in the germ line. Genetic testing for the presence of this specific mutation in siblings or offspring should be pursued. These children can then be aggressively surveyed for the presence of emerging tumors. If genetic testing is not pursued, then tumor surveillance is recommended for all siblings of the affected patient. Current recommendations suggest examination at birth and every 4 months for up to 4 years of age. For children with unilateral disease, genetic screening for RB1 mutations can now be offered to families at the time of enucleation. The testing requires a sample of tumor and peripheral blood from the patient,37 or, for patients with bilateral disease, a blood sample can be analyzed directly.
CLINICAL PRESENTATION
Most cases of retinoblastoma in the United States are diagnosed while the tumor remains intraocular without local invasion or distant metastases. In developing countries, however, the diagnosis is frequently made only after an enlarged eye or gross orbital extension is apparent. These patients more commonly present with local invasion.
The signs and symptoms of an intraocular tumor depend on its size and position. The most common presenting sign is leukocoria of one or both eyes (Fig. 28.1). Leukocoria, a lack of the normal red reflex of the eye, is manifest when the tumor is large or has caused a total retinal detachment leading to a retrolental mass that is visible through the pupil. If vitreous hemorrhage occurs because of bleeding from the retinoblastoma vessels, the pupil may appear to have a dark reflex instead of the white reflex typically seen in retinoblastoma18,74,75 (Fig. 28.2). The second most common presenting sign is strabismus. Loss of central vision from a
P.868

tumor in the macula may result in a disruption of the fusional reflex and cause the affected eye to drift.
Figure 28.1 Photograph of an eye from a patient with retinoblastoma who presented with leukocoria.
Figure 28.2 A: Gross photograph of an eye with retinoblastoma (white membranous tissue at center of the eye) with subretinal tumor seeds (arrow), and vitreous and subretinal hemorrhages (H). Neovascularization of the anterior chamber and partial closure of the anterior angle (*) are also present. B: Histologic picture of rubeosis iridis showing neovascularization (arrows) of the anterior portion of the iris (i) and focally on the endothelial surface of the cornea (arrow). Contraction of the neovascular membrane produces closure of the anterior chamber angle (*) (hematoxylin and eosin, original magnification 20×). C: Histologic picture of the anterior segment of an eye with retinoblastoma seeds on the surface of the iris (i) with focal rosette formation (insert). The anterior chamber angle (*) is opened (hematoxylin and eosin, original magnification 20×). (See Color Figure 28.2.)
Other ophthalmic features accompany some cases of retinoblastoma and may indicate the necessity for immediate enucleation. Heterochromia (different color for each iris) may present as an initial sign of retinoblastoma secondary to iris neovascularization. The diagnosis of retinoblastoma should be excluded in children who present with this condition.20 Rubeosis iridis (neovascularization of the surface of the iris) occurs in approximately 17% of patients with retinoblastoma and in more than 50% of patients with advanced retinoblastoma requiring enucleation20,76,77,78,79 (Fig. 28.2). Extensive necrosis of the tumor and liberated angiogenic factors may be responsible for this neovascularization of the iris.
Spontaneous bleeding from rubeosis iridis may also cause hyphema (blood in the anterior chamber), and the potential diagnosis of retinoblastoma should be investigated in a child presenting with spontaneous hyphema without history of trauma.18,20,74,75 Glaucoma may be secondary to neovascularization of the anterior chamber angle and/or anterior synechia as a result of rubeosis iridis. Closed-angle glaucoma can also be secondary to mechanical obstruction of the anterior chamber angle by the iris and lens that has been pushed forward by a large intravitreal tumor. Most children with these presentations undergo enucleation.18,74,75 Anterior chamber seedings from endophytic tumors or diffuse infiltrating tumors may produce pseudohypopyon (cells in the anterior chamber) (Fig. 28.2). Intraocular tumors are not associated with pain unless secondary glaucoma or inflammation is present.
DIAGNOSIS
Most commonly, a parent or relative of an affected child notes an abnormality of the eye that prompts physician evaluation. Current detection strategies involve a pediatrician looking for leukocoria using an ophthalmoscope. The gross appearance of a creamy pink to snow white mass projecting into the vitreous during the ophthalmoscopic examination (Fig. 28.1) may suggest retinoblastoma; however, associated findings of retinal detachment, vitreous hemorrhage, or opaque media often make inspection difficult. Pupillary dilation and examination with the patient under anesthesia are essential to fully evaluate the retina. Characteristically, the diagnosis is made by the ophthalmoscopic, radiographic, and ultrasonographic appearance and pathologic confirmation is unnecessary. When the tumor is in an advanced stage, distinguishing vitreal seeding from multifocal tumors may be difficult; however, this distinction has important ramifications for the prognosis for the patient and for genetic counseling for the family. Earlier detection of the tumor would benefit the patient both by decreasing the chance of a child presenting with metastatic disease and by increasing the chance of being able to salvage the affected eye. A suggestion has been made to include dilation of the pupil prior to examination at the first well-child visit. An additional benefit of screening would be the earlier detection and treatment of congenital and infantile cataracts. Whether routine screening would be practical is controversial because diseases such as retinoblastoma and congenital cataracts are rare (congenital cataracts affect approximately 1 in 2,000 live births) and because pediatricians may not be adequately trained to recognize these conditions.
Ultrasonography and computed tomography (CT) (Fig. 28.3) of the orbit are the imaging studies most frequently used to confirm the diagnosis of retinoblastoma and to detect ectopic disease in the pineal gland.80 Magnetic resonance imaging (MRI) of the orbit (Fig. 28.3) may be a more useful technique to detect tumor extension into the optic nerve and orbital coats.81
The pretreatment evaluation must be individualized for each patient. In patients who present with small tumors, ultrasonography and/or CT scan of the orbits and careful examination under anesthesia may be all that is necessary to make the diagnosis. A more extensive metastatic workup is not necessary in these patients unless there is a question of optic nerve extension or extensive choroidal invasion. A lumbar puncture to obtain
P.869

cerebrospinal fluid (CSF) for cytologic study and CT or MRI of the brain to rule out brain metastases can be performed in those patients. Because of the rarity of distant metastases in patients with retinoblastoma, a bone marrow examination or bone scan is usually not warranted unless the physician has suspicions of systemic involvement.
Figure 28.3 Diagnostic imaging of children with retinoblastoma. A: CT scan of the orbit. Axial view showing a partially calcified mass (T) consistent with retinoblastoma, a normal lens (L) and normal optic nerve (ON). The globe is intact and shows no evidence of extraocular invasion by tumor. B: MRI of the orbits and brain. Contrast-enhanced coronal T1-weighted image showing parasellar and left middle fossa spread of retinoblastoma (T) with extension along the sylvian fissure. The eye on the left is normal. The orbit on the right contains a prosthesis.
Differential Diagnosis
A number of benign conditions can clinically simulate retinoblastoma (pseudoretinoblastomas) and sometimes create considerable diagnostic difficulty for the ophthalmologist. Clinical definition is mandatory because the management of these entities differs considerably from the radical treatment of retinoblastoma.
Early reports of the frequency of enucleations performed for suspected retinoblastomas when an alternative final pathologic diagnosis is made varied from 30% to 16% according to the degree of oncologic experience of and the type of referrals received by the group reporting the series.18,75,82,83 Most clinicians are now more familiar with pseudoretinoblastomas, and the frequency of erroneous enucleation is currently much lower.20
Other conditions that might be confused clinically also produce or simulate a mass in the vitreous or the retina. With the exception of medulloepithelioma, these lesions have in common a variety of histopathologic features distinct from retinoblastoma that create a difficult differential diagnosis for the pathologist.77,84,85,86,87,88,89,90,91,92,93,94,95,96
Approximately 60% of pseudoretinoblastomas include the differential diagnosis of three nonneoplastic entities: Toxocara canis endophthalmitis, persistent hyperplastic primary vitreous (PHPV), and Coats’ disease.18 All of these entities might present with retinal detachment and may have retrolenticular fibrosis. T. canis endophthalmitis is caused by the larvae of the nematode T. canis and presents almost always in children, although never at birth. Clinical history and serology are important for the diagnosis. Usually there are no signs of ocular inflammation, because the live larvae do not elicit an inflammatory response. Dead larvae elicit the formation of a localized eosinophilic abscess surrounding the microorganism. Condensed vitreous with gliosis and fibrosis may be present at the site of infection. Because these organisms are very small and degenerate, histologic confirmation is very difficult.
PHPV is a congenital anomaly of the primary vitreous in which embryonal vessels do not regress and may pull the retina, resulting in an anterior detachment. If the posterior capsule of the lens is ruptured by the traction of the vessels and fibrous membrane, then a posterior subcapsular cataract forms. Some cases of PHPV are associated with a wide band to the optic nerve and with retinal dysplasia. There are rare cases reported when PHPV has been associated with retinoblastoma.97,98,99,100,101
In contrast to toxocariasis and PHPV, Coats’ disease lacks the fibrosis and vascularization of vitreous. Coats’ disease is characterized by peripheral retinal vascular telangiectasis. These abnormal vessels leak and create an exudative retinal detachment rich in lipids with subretinal foamy macrophages and cholesterol clefts. Toxocariasis and PHPV simulate endophytic retinoblastoma, and Coats’ disease mimics the exophytic type (see the section on pathology, gross features).82,86,102,103,104,105,106 Ancillary imaging technology that includes ultrasound, CT scans, and MRI have greatly helped in the differential diagnosis of these lesions.97,106,107,108,109,110,111
Use of Cytology in the Diagnosis of Retinoblastoma
Retinoblastoma is the most frequent intraocular tumor in childhood in the United States, and in some other countries retinoblastoma is the most frequent intraocular tumor overall. However, retinoblastoma is one of the only human tumors that are radically treated without tissue biopsy confirmation. The clinical presentation and the ancillary radiologic and ultrasonic findings are typical for retinoblastoma in the majority of patients. Usually the correct diagnosis does not represent a diagnostic dilemma for the experienced pediatric/oncologic ophthalmologist. The resistance to biopsy confirmation of the tumor arises from the dramatic difference in survival for patients with contained intraocular tumors versus those with extraocular seeding of the tumor. The bias against biopsy is also aggravated by the reports of cases in which the tumor was misdiagnosed as “uveitis” or obscured by cataract and patients developed orbital extensions of retinoblastoma after vitrectomy.112 Aggressive postvitrectomy therapy has resulted in prevention of metastasis in patients with unsuspected retinoblastoma.113
Figure 28.4 A: Drawing of an eye with retinoblastoma indicating the entrance of the needle through the peripheral cornea and peripheral iris, then between the ciliary body and the lens into the tumor. This technique prevents the spreading of tumor cells via the needle tract by avoiding the vascularized conjunctiva, orbit, sclera, and pars plana. B: Cytologic preparation of a retinoblastoma showing cohesive groups of neoplastic cells with high nuclear to cytoplasmic ratio, increased mitotic activity (dotted arrow), and focal rosette formation (solid arrow) (Papanicolaou stain, original magnification 100×). C: Cytologic preparation of cerebrospinal fluid (CSF) in a patient with retinoblastoma metastatic to the brain. Notice the cohesive groups of neoplastic cells with high nuclear to cytoplasmic ratio (hematoxylin and eosin stain, original magnification 20×).
P.870

Recently, the development of more refined techniques of fine-needle aspiration biopsy (FNAB) and the increased knowledge of the biologic behavior of retinoblastoma have allowed some patients to benefit from pretreatment biopsy.102,114,115,116,117,118,119,120,121 This FNAB technique is more difficult to perform. However, FNAB is safer for the patient because it prevents tumor seedings by avoiding the subconjunctival and scleral/orbital routes of entry. The 30-gauge needle passes through the peripheral cornea, anterior chamber, peripheral iris, lens zonules (avoiding puncture of the lens), and vitreous and penetrates the tumor (Fig. 28.4). There have been no reports of extraocular tumor spread through the needle tract when FNAB has been used. FNAB is recommended only in selected cases in which the diagnosis is ambiguous and when adequate steps are taken to prevent extraocular seeding of tumor cells.122
The cytologic findings are those of small to medium-sized basophilic cells with scanty cytoplasm that tend to group together (rosette-like). Mitoses may be easy to find and necrosis is frequently encountered (Fig. 28.4). Similar features can be seen in cytologic CSF specimens in children with intracranial metastases (Fig. 28.4). In one series of FNAB of pediatric intraocular tumors, the overall accuracy of FNAB was 95%, and the accuracy of cytologic interpretation was 100%. Therefore, FNAB is a reliable and accurate diagnostic tool for the assessment of selected pediatric ophthalmic diseases when the diagnosis is in question.123,124
PATHOLOGY
Gross Features
Primary retinoblastomas originate in the sensory retina and occupy the retina and vitreal cavity. Retinoblastoma is usually white-gray with a chalky appearance and a soft, friable consistency. Bright white speckles corresponding to calcifications present throughout the tumor. The gross features of retinoblastoma depend on the growth pattern of the tumor.18,74,125 Some of these patterns correlate with clinical presentations and differences in biologic behavior, especially as they relate to intraocular and extraocular types of tumor spread.
The endophytic growth pattern is represented by tumors arising from the retina and growing into the vitreal cavity (Fig. 28.5). These tumors tend to entirely fill the cavity and produce floating tumor spheres called vitreal seeds. Tumor left
P.871

untreated eventually invades the anterior portion of the eye, reaching the aqueous venous channels and the conjunctiva. From there, the tumor can permeate the lymphatic vessels and metastasize to regional lymph nodes.18,125,126,127,128
Figure 28.5 A: Gross photograph of an eye with a retinoblastoma showing an endophytic growth pattern. Notice that the tumor mass is growing from the retina (arrow) into the vitreal cavity. B: Gross photograph of an eye with a retinoblastoma showing an exophytic growth pattern. Notice that the tumor is growing from the retina (arrow) into the subretinal space with associated retinal detachment. C: Gross photograph of an eye with a retinoblastoma showing a mixed growth pattern, the most frequent type. Notice that the tumor grows both into the vitreous cavity and into the subretinal space with the retina (arrow) entrapped in the middle. The tumor has massively invaded the choroid (C). D: Gross photograph of an eye with a diffuse retinoblastoma. Notice the absence of a well-formed mass; instead there are white seeds of tumor cells along the retina (arrow) and ciliary body (cb). (See Color Figure 28.5.)
Exophytic tumors grow from the retina into the subretinal space and often cause serous detachments of the retina (Fig. 28.5). These tumors may invade the choroid through Bruch’s membrane.18,125,129,130 Mixed endophytic and exophytic tumor growth is the most common pattern encountered18,20,125 (Fig. 28.5).
Diffuse infiltrating retinoblastoma is the least common tumor growth pattern but is the most diagnostically challenging type because there is no predominant mass (Fig. 28.5). This presentation of retinoblastoma is seen in children with an average age of 6 years. The tumor cells grow throughout the retina while single cells and vitreous seeds invade the anterior portions of the retina, the ciliary body, and eventually the anterior chamber. Clinically this type of tumor resembles an inflammatory process with pseudohypopyon mimicking inflammatory cell accumulation (hypopyon) and vitreous seeds simulating the inflammatory cellular reaction seen in uveitis. Because this type of retinoblastoma resembles an inflammatory process, the diagnosis is often delayed until cytologic examination of the aqueous humor or, in rare cases, of the vitreous. Almost all reported cases have a unilateral, sporadic presentation without family history. Although the diagnosis is difficult, children with diffuse infiltrating retinoblastoma have a good prognosis after enucleation.99,128,131,132,133,134,135,136,137,138,139,140 Any child, regardless of age, who presents with signs of endophthalmitis should be considered to have diffuse infiltrating retinoblastoma until proven otherwise.138
Complete spontaneous tumor regression through unknown mechanisms occurs more commonly in retinoblastoma than in any other malignant tumor. In most of these cases, complete occlusion of the central retinal artery is found; however, it is not known whether this is a primary event or the result of tumor necrosis.141,142 Severe inflammatory reaction with massive necrosis of the tumor followed by phthisis bulbi (complete atrophy of the eye) is the usual presentation.141,142,143,144,145,146,147,148,149,150,151,152,153 If the eye is enucleated at the time of acute necrosis, the gross findings are those of massive tumor necrosis with edema of the conjunctiva.141,154 If the eye is examined after complete atrophy has occurred, the findings are those of a small shrunken eye with mostly necrotic calcified tumor and a disorganized retina.
Histologic Features
Microscopic examination of the affected eye displays one or more tumors with large areas of necrosis and multifocal calcifications replacing portions of the retina (Fig. 28.6). The majority of the tumor is formed by small hyperchromatic cells with a high nuclear to cytoplasmic ratio. The tumor cells are mitotically active but frequently exhibit apoptosis18,20,74,93,125,155,156,157,158,159 (Fig. 28.6). The viable cells surround blood vessels in a range of 90 to 110 mm forming a collarette (pseudorosettes) (Fig. 28.6). Viability of the tumor cells depends on the intrinsic tumor blood supply. Areas of coagulative necrosis contain multiple foci of dystrophic calcification. Tumor cell necrosis liberates DNA from the nuclei of the cells. The released DNA forms deposits on the basement membranes of the vessels, the lens (capsule), the retina (internal limiting membrane), and the choroid (Bruch’s membrane)160 (Fig. 28.6).
Some retinoblastomas show large areas of undifferentiated or poorly differentiated tumor (Fig. 28.7); other retinoblastomas show a certain degree of differentiation represented by formation of rosettes. Flexner-Wintersteiner rosettes are highly characteristic of retinoblastoma, although they are also seen in pinealoblastomas and medulloepitheliomas. Flexner-Wintersteiner rosettes are lined by tall cuboidal cells that circumscribe an apical lumen. The apical ends attach to each other by terminal bars, and the cells may have apical cytoplasmic projections into the lumen of the rosette (Fig. 28.7). Electron microscopy has demonstrated that these projections represent inner and outer segments of photoreceptors.91,161,162 This and several other observations support the idea that retinoblastomas arise from undifferentiated retinal cells that may differentiate into photoreceptors91,161,162,163,164,165,166,167,168 usually of the cone cell lineage.6,7 An explanation for this observation may have been found in the mouse where RB expression is required for retinal progenitor cells to exit the cell cycle and for rod photoreceptor differentiation to occur.169
P.872

Homer Wright rosettes are less common than Flexner-Wintersteiner rosettes, and they are found in a variety of neuroblastic tumors in addition to retinoblastoma. These rosettes do not surround a lumen but rather extend cytoplasmic processes that fill the center of the rosette. Homer Wright rosettes may be incomplete and admixed with well-formed Flexner-Wintersteiner rosettes (Fig. 28.7).
Figure 28.6 A: Histologic picture of a retinoblastoma growing from the retina (R) into the vitreous cavity (V) (hematoxylin and eosin stain, original magnification 4×). B: Histologic photograph showing poorly differentiated retinoblastoma cells with high nuclear to cytoplasmic ratio, increased mitotic activity (solid arrow), and increased apoptosis (hollow arrow) (hematoxylin and eosin stain, original magnification 100×). C: Microphotograph showing viable tumor cells surrounding blood vessels (bv) with necrotic cells beyond a rim of approximately 110μm (hematoxylin and eosin stain, original magnification 20×). D: Microphotograph of retinoblastoma with extensive necrosis (insert) with foci of calcification (*). Notice the large vessel darkly stained with hematoxylin (basophilic) secondary to DNA deposits in the vascular basement membrane as a result of extensive tumor cell necrosis that liberates nuclear DNA (hematoxylin and eosin stain, original magnification 10×).
Figure 28.7 A: Microphotograph of a poorly differentiated retinoblastoma showing sheets of neoplastic cells without rosette formation (hematoxylin and eosin stain, original magnification 40×). B: Microphotograph of a retinoblastoma showing Flexner-Wintersteiner rosette formation (insert). Notice that these rosettes have a center partially filled by cytoplasmic prolongations with apical terminal bars (hematoxylin and eosin stain, original magnification 40×). C: Microphotograph of a retinoblastoma showing Homer Wright rosette formation (insert). The lumen of the rosette is filled by cytoplasmic prolongations (hematoxylin and eosin stain, original magnification 40×). D: Microphotograph of a well-differentiated retinoblastoma showing fleurettes (insert). Fleurettes are groups of well-differentiated cells similar to photoreceptors joined by cytoplasmic junctions and forming a figure similar to a bouquet of flowers (hematoxylin and eosin stain, original magnification 40×).
About 6% of tumors show benign photoreceptor differentiation into groups of cells with short cytoplasmic processes, abundant cytoplasm, and small round nuclei similar to photoreceptors. These groups of cells, which resemble a bouquet of flowers, are called fleurettes.16,170,171 Neither significant mitotic activity nor necrosis is observed within the fleurettes172,173,174 (Fig. 28.7).
Retinocytoma
A benign counterpart of retinoblastoma called retinocytoma (also retinoma) that solely contains well-differentiated glial cells and fleurettes has recently been described. These benign tumors contain areas of abrupt calcification associated with retinal pigment epithelium proliferation (Fig. 28.8). They exhibit specific features that allow experienced clinicians to follow the behavior of the tumor without radical treatment.4,9,74,175,176 Singh et al.175 described the ophthalmoscopic features of 24 tumors considered to be characteristic of retinocytoma, including the presence of a translucent retinal mass in 21 (88%), calcification in 15 (63%), and retinal pigment epithelial alteration in 13 (54%) of the tumors. A combination of all three features was observed in 8 (33%) of the 24 tumors. In 13 (54%) of the tumors, a zone of chorioretinal atrophy could be observed. Although the majority of these tumors behave as benign lesions, close follow-up is suggested because a few tumors have been reported to have undergone malignant transformation into retinoblastomas that eventually required enucleation.170,175
Figure 28.8 A: Histologic picture of abrupt cell calcification in a retinocytoma (hematoxylin and eosin stain, original magnification 40×). B: Histologic picture of a well-differentiated area of a retinocytoma showing glial and neural differentiation (*) (hematoxylin and eosin stain, original magnification 40×). C: Histologic picture of a well-differentiated area of a retinocytoma showing fleurettes (arrows) (hematoxylin and eosin stain, original magnification 40×).
In retinocytoma tumors that have undergone complete regression, either spontaneously or secondary to treatment,
P.873

mummified calcified tumor cells and large areas of dystrophic calcification are observed. Exuberant reactive retinal pigment epithelial proliferation, ciliary epithelial cells, and glial cells with occasional ossification accompany this process.74
METASTASIS AND RECURRENCE
If left untreated, retinoblastoma usually fills the eye and completely destroys the internal architecture of the globe. The most common route of spread is by invasion through the optic nerve. Once in the nerve, tumor spreads directly along the nerve fiber bundles toward the optic chiasm or infiltrates through the pia into the subarachnoid space. From the subarachnoid space, the retinoblastoma can involve the CSF, the brain, and the spine. The second major route of spread is through massive involvement of the choroid into the orbit either via scleral canals (areas within the sclera where ciliary vessels, nerves, and vortex veins enter or exit the eye) or by direct extension through the sclera.177 Extraocular extension generally occurs within 6 months if intraocular tumors are left untreated.
Extraocular extension dramatically increases the chances of hematogenous and lymphatic spread. There are four routes for metastatic spread of retinoblastoma.18,74
Metastatic spread can occur by direct infiltration either through the optic nerve into the brain or through the choroid into the orbit soft tissues and bones.
Dispersion of the tumor cells through the subarachnoid space of the optic nerve into the opposite optic nerve or through the CSF into the brain and spine can cause metastatic spread of the tumor. This can occur without detectable presence of retinoblastoma at the surgical margin of the optic nerve.
  • Metastasis can occur by hematogenous dissemination secondary to orbital and bone invasion or when lymphatic invasion reaches the lymph nodes. Among other sites, widespread metastasis can present in lung, bone, and brain.
  • Metastasis via lymphatic dissemination occurs in tumors that spread anteriorly into the conjunctiva and eyelids or extend into extraocular tissues. Lymphatic vessels and lymphoid tissue are absent in the orbit and intraocular tissues. In the ocular region, only conjunctiva and skin have lymphatic channels. Tumors must first reach these areas to permeate the lymphatic vessels and then spread into regional lymph nodes.
Histologically, retinoblastoma metastases appear less differentiated than intraocular tumors. Rosettes are rarely encountered and fleurettes have never been described. When very well differentiated extraocular tumors appear outside the orbit, a differential diagnosis of a primary primitive neuroectodermal tumor (PNET) must be considered.
TRILATERAL RETINOBLASTOMA AND OTHER TUMORS
Primary retinoblastomas of the pineal and parasellar sites have been called trilateral retinoblastomas and usually present as single tumors. Trilateral retinoblastoma is a well-recognized, although rare, syndrome.178 The majority of the reported cases have involved patients with a family history of retinoblastoma, and the disease is usually fatal. These tumors may appear several years after successful treatment of intraocular retinoblastoma. They may be far more differentiated than the primary tumor and may contain numerous rosettes, fleurettes, and individual cells showing photoreceptor differentiation. The presentation of trilateral retinoblastoma contrasts with metastatic retinoblastoma because metastatic retinoblastoma presents as multiple, undifferentiated tumors within the first 2 years of initial treatment.
STAGING
Different classifications have been introduced as guidelines for predicting prognosis for vision, for globe salvage, and for life. The Reese-Ellsworth classification179 (Table 28.1) relates to eyes treated by methods other than enucleation, specifically by radiotherapy. This classification, devised prior to the development of current ophthalmologic methodologies for diagnosis and treatment, has often been used to imply prognosis for life rather than for vision. Although the Reese-Ellsworth classification is not necessarily prognostic for outcomes using modern treatment modalities, it is still the classification used most often to compare therapeutic results. The American Joint Committee on Cancer (AJCC) has proposed a clinical and pathologic staging classification for retinoblastoma in which complete spontaneous regression of the tumor has not occurred. Using these criteria for cases of bilateral retinoblastoma, each eye is staged separately. Histologic verification of the disease in an enucleated eye is required, and any unconfirmed cases must be reported separately. The extent of retinal involvement is indicated as a percentage of the total retinal area. For pathologic staging, all of the clinical and pathologic data from the resected specimen are to be used. A revised staging classification based on clinical presentation may better predict the success of globe salvage using current treatment modalities. One such staging system has been proposed180 (Table 28.2).
Prognostic Factors
Prognosis for vision in children with unilateral retinoblastoma is excellent for the uninvolved eye. The development of tumors in the contralateral eye after 3 years is very rare. A primary tumor with vitreous, subretinal, and retinal seeds can be mistaken as a multifocal primary tumor. Primary tumors arise from the sensory retina in contrast to retinal seedings, which sit on top of the retina (Fig. 28.9) or on the inner subretinal surface of the photoreceptors. The presence of multiple primary tumors or the emergence of tumors in both eyes (bilateral retinoblastoma) supports a diagnosis of inherited retinoblastoma. The prognosis for vision in bilateral retinoblastoma depends on the extent of tumor involvement and the effectiveness of treatment modalities. If the tumors are small and away from the fovea (central portion of the retina
P.874

with best visual acuity), one may anticipate a good prognosis for vision after successful treatment.18,74,125
TABLE 28.1 REESE-ELLSWORTH CLASSIFICATION FOR CONSERVATIVE TREATMENT OF RETINOBLASTOMAa
Group I Very favorable
  a Solitary tumor, less than 4 disc diameters in size, at or behind the equator
b Multiple tumors, none over 4 disc diameters in size, all at or behind the equator
Group II Favorable
a Solitary tumor, 4 to 10 disc diameters in size, at or behind the equator
b Multiple tumors, 4 to 10 disc diameters in size, behind the equator
Group III Doubtful
a Any lesion anterior to the equator
b Solitary tumors larger than 10 disc diameters behind the equator
Group IV Unfavorable
a Multiple tumors, some larger than 10 disc diameters
b Any lesion extending anteriorly to the ora serrata
Group V Very unfavorable
a Massive tumors involving over half the retina
b Vitreous seeding
aRefers to chances of salvaging the affected eye and not systemic prognosis.
The survival rate from retinoblastoma has improved dramatically over the last century. One of the first retinoblastoma survival studies was reported in 1897 by Wintersteiner.180 That 13% survival rate is in sharp contrast to the 90% overall 5-year survival reported by many centers today.181,182,183,184 The main reasons for this improvement are the improved ability to detect retinoblastoma prior to the onset of metastatic disease and the development of alternative treatment strategies (see the section on therapeutic options).
TABLE 28.2 PRACTICAL GROUPING METHOD OF OCULAR RETINOBLASTOMA BASED ON GENERAL CLINICAL FEATURES
Group Abbreviation Clinical Features
1 T Tumor onlya
2 T+SRF Tumor plus subretinal fluid
3 T+FS Tumor plus focal seeds
a. SRS ≤3 mm from tumor
b. VS ≤3 mm from tumor
4 T+DS Tumor plus diffuse seeds
a. SRS >3 mm from tumor
b. VS >3 mm from tumor
5 High risk Tumor plus (any one)
Neovascular glaucoma
Opaque media from hemorrhage in anterior chamber, vitreous, or subretinal space
Invasion of postlaminar optic nerve, choroid (>2 mm), sclera, orbit, anterior chamber
DS, diffuse seeds; FS, focal seeds; SRF, subretinal fluid; SRS, subretinal seeds; T, tumor, VS, vitreous seeds.
aRegardless of tumor size.
Modified from Shields CL, Mashayekhi A, Demirci H, et al. Practical approach to management of retinoblastoma. Arch Ophthalmol 2004;122:729–735.
Metastatic disease is still associated with a poor prognosis. Most clinical findings are not useful in predicting the occurrence of metastasis in children with retinoblastoma, although histopathologic data provide a fair estimate of its risk. Multivariate statistical analysis has suggested the correlation of certain histopathologic findings and prognostic risk factors.183,185,186,187,188,189 The most important prognostic indicators for the development of metastasis are the presence of tumor in the optic nerve posterior to the lamina cribrosa at the site of surgical transection and extrascleral extension of tumor into the orbit.102,126,127,129,177,183,186,189,190,191,192,193,194,195,196 The extent of
P.875

tumor invasion in the optic nerve correlates with prognosis (Fig. 28.10). Superficial invasion of the optic disc is associated with a mortality rate of 10%, a rate similar to that seen when the optic nerve is not involved. The presence of tumor up to the lamina cribrosa is associated with a mortality rate of 29%. Invasion of tumor posterior to the lamina cribrosa is associated with a mortality rate of 42%, and the presence of tumor at the transected surgical margin is associated with a mortality of 80%.126,177,183,186,189,197 The importance of obtaining a large portion of optic nerve at the time of enucleation is underscored by these results. Specific studies related to the length of the optic nerve stump alone show that patients with the optic nerve measuring less than 5 mm in the enucleated eye have a worse prognosis than those having stumps greater than 5 mm.183,193,198,199,200
Figure 28.9 A: Gross photograph showing a primary retinoblastoma tumor (T) with tumor seeds (ts) on the inner retina (arrow) and in the vitreous (v). B: Microphotograph of an area of vitreous tumor seeds. Notice the hollow center of the retinoblastoma seeds. The retina (arrow) is without tumor (hematoxylin and eosin stain, original magnification 10×). C: Microphotograph of a retinoblastoma seed on the inner surface of the retina. Notice the rosette formation of the tumor and the intact architecture of the retina (arrow) (hematoxylin and eosin stain, original magnification 40×). D: Microphotograph of a retinoblastoma arising from the retina (arrow). Notice the tumor replacing the normal architecture of the retina (hematoxylin and eosin stain, original magnification 20×). (See Color Figure 28.9.)
Massive, but not focal, invasion of the choroid by tumor increases the possibility for hematogenous spread, either through vascular permeation of choroidal vessels or more frequently by extension through the sclera into the orbital tissues126,183,188,190 (Fig. 28.10). MRI studies are helpful in evaluating the extent of involvement of choroid or optic nerve by tumor.111
Retinoblastomas that are poorly differentiated tend to behave more aggressively and are associated with a worse prognosis. Other factors associated with some risk for metastatic behavior, especially in conjunction with the major factors cited previously, are tumor invasion into the anterior chamber, large tumor size with vitreous seeding, rubeosis iridis, and glaucoma.
THERAPEUTIC OPTIONS
The management of retinoblastoma is complex. The diagnosis and treatment of patients with retinoblastoma involves a team approach requiring pediatric oncologists, ophthalmologists, and radiologists skilled in the treatment of patients with retinoblastoma. Important team roles are
P.876

also filled by child psychologists, social workers, nurses, and genetic counselors who can support families dealing with the difficulties of caring for a child who not only has cancer but also may lose an eye and vision. The goals of treatment are, most importantly, to save the child’s life and, second, to salvage the eye and/or vision. Therapy is tailored to the individual case and is based on the overall situation, including threat of metastatic disease, risks for second cancers, systemic status, laterality of the disease, size and location of the tumor(s), and visual prognosis. There are several medical and surgical options for treatment of retinoblastoma, and the ocular oncologist should be thoroughly familiar with the indications, techniques, and expected outcomes as well as the associated systemic and visual problems of all treatment methods.201 The currently available treatment methods for retinoblastoma include enucleation, external beam radiotherapy, plaque radiotherapy, laser photocoagulation, cryotherapy, thermotherapy, chemothermotherapy, intravenous chemoreduction, subconjunctival chemoreduction, systemic chemotherapy for possible metastatic disease, and orbital exenteration.
Figure 28.10 A: Microphotograph of a normal optic nerve showing the prognostic percentages of survival of patients with invasion of the optic nerve by anatomic portions of the nerve. Patients with tumors invading the pre-lamina cribrosa have a 10% mortality rate similar to that seen without invasion of the nerve. Invasion into the lamina cribrosa (LC—between the two dotted lines) carries a 29% mortality rate, and invasion beyond the lamina cribrosa carries a 42% mortality rate. Patients with tumors that are present at the surgical margin of resection (single dotted line) have an 80% mortality rate (myelin stain, original magnification 10×). B: Microphotograph of an optic nerve showing tumor invasion beyond the lamina cribrosa (LC) but not at the surgical margin of resection (hematoxylin and eosin stain, original magnification 10×). C: Microphotograph of the posterior pole and optic nerve of an eye with massive involvement of the choroid (C) and optic nerve by tumor. Notice that retinoblastoma tumor is present at the surgical margin of resection of the optic nerve (single dotted line) (hematoxylin and eosin stain, original magnification 4×). D: Microphotograph of the subretinal space (SR) and choroid (C) of an eye with retinoblastoma with focal involvement of the subretinal space and minimal involvement of the choroid (hematoxylin and eosin stain, original magnification 20×).
Enucleation
Enucleation is still the treatment of choice for advanced retinoblastoma with concern of tumor invasion into the optic nerve, choroid, or orbit and no hope for salvage of useful vision in the affected eye. Those eyes with secondary glaucoma, pars plana seeding, or anterior chamber invasion are also generally best managed with enucleation.
In the past, most children with unilateral retinoblastoma were managed with enucleation. Those patients with bilateral retinoblastoma usually had the most advanced eye enucleated and the less advanced eye treated with external beam radiotherapy.20 This management philosophy has been gradually modified with the advent of newer, more conservative but effective methods.200,202,203 There has been a substantial decrease in the frequency of enucleation over recent decades.204 In a review of 324 consecutive cases of retinoblastoma managed on the Oncology Service at Wills Eye Hospital from 1974 to 1988, Shields and coworkers204 found that unilateral retinoblastoma was managed with enucleation in 96% of cases from 1974 to 1978, in 86% of cases from 1979 to 1983, and in 75% of cases from 1984 to 1988. A similar decreasing trend was found with bilateral retinoblastoma.204 The frequency of enucleation is even less today.
Enucleation involves the gentle removal of the intact eye without seeding the malignancy into the orbit. Special care must be taken to perform all steps in a controlled fashion to avoid globe perforation or compression. Shields et al. have described the surgical technique for enucleation of an eye with retinoblastoma.20,205,206 Because the underlying sclera is thin at the site of muscle insertions, the rectus muscles are handled delicately when the hook is placed flat along the sclera. At the time of optic nerve cutting, scleral or muscle insertion traction sutures are avoided to prevent inadvertent globe perforation. Mild traction with a hemostat on the medial rectus muscle stump is used to lift the globe cautiously to avoid inadvertent lamellar rip of the sclera and cornea, which could threaten the integrity of the eye. Optic nerve snares or clamps should be avoided because they induce more vigorous trauma to the eye and can produce crush artifact in the optic nerve. This artifact can cause difficulty for the pathologist assessing the possibility of retinoblastoma invasion of the optic nerve. The use of minimally curved enucleation scissors is preferred to achieve a long optic nerve section (Fig. 28.11).
Figure 28.11 Enucleation and fresh tissue harvesting. A long section of optic nerve is obtained with the globe at enucleation of an eye with retinoblastoma. The posterior aspect of the optic nerve is cut and submitted to pathology separately for analysis of optic nerve invasion.
Historically, an orbital implant was not usually placed after enucleation for retinoblastoma because of potential interference with palpation of the socket and clinical detection of orbital tumor recurrence. More recently, with improved knowledge of the behavior of retinoblastoma and the risks of local orbital recurrence, there is less hesitation for placing an orbital implant. Available orbital imaging modalities, including CT and MRI, allow detailed orbital analysis despite the presence of an implant. The orbital implant provides a more natural cosmetic appearance of the patient’s artificial eye, minimizes sinking of the prosthesis, and enables motility of the prosthesis. Orbital implants made of polymethylmethacrylate sphere, coralline hydroxyapatite, bovine hydroxyapatite, or polyethylene are commonly used.205,207 A tissue wrap is usually provided to these implants so that the four rectus muscles can be anatomically reattached to the implant and provide implant motility with little resistance in the orbit. Available tissue wraps are many and include povidone-iodine–treated human sclera, irradiated human sclera, bovine pericardium, fascia lata, and Vicryl mesh, among others. The motility implant, when properly placed surgically, has been shown to be well tolerated by children and adults.205,206,208
External Beam Radiotherapy
Retinoblastoma is generally a radiosensitive tumor. External beam radiotherapy is a method of delivering whole-eye irradiation to treat advanced retinoblastoma, particularly when there is diffuse vitreous seeding. The whole-eye and lens-sparing techniques used currently have been shown to improve the eye preservation rate as compared to reported older techniques. The rate of ocular salvage depends on the Reese-Ellsworth stage of the disease at the time of treatment as well as on the availability of focal therapy for limited
P.877

recurrences.209,210 Recurrence of retinoblastoma after external beam radiotherapy continues to be a problem and can develop within the first 1 to 4 years after treatment.211 Tumor recurrence in other studies has also been found to be related to the stage of the disease and to the size of the largest tumor at the time of treatment.211,212,213,214 Prophylactic radiotherapy to a normal contralateral eye is almost never indicated today.215
Little has been written on the visual outcome after external beam radiotherapy for retinoblastoma. Radiation damage to the retina, optic nerve, and lens can be challenging to manage.216 Patients with macular retinoblastoma have visual outcomes that are dependent on the size of the tumor and the degree of involvement of the fovea.217 Superimposed amblyopia can pose a problem, and patching therapy should be used if hope for vision remains.
External beam radiotherapy may induce a second cancer in the field of irradiation. The 30-year cumulative incidence for second cancers in bilateral retinoblastoma has been reported to be 35% for patients who received radiation therapy compared with 6% for those who did not receive radiation.218 Overall, the cumulative probability of death from second primary neoplasms was reported at 26% at 40 years after bilateral retinoblastoma diagnosis. External beam radiotherapy has been reported to further increase the risk of mortality from second neoplasms.40 Abramson and Frank219 found that external beam radiotherapy increased the incidence of second cancers in the field of radiation but did not stimulate second cancers outside the field of irradiation. In their series, patients younger than 12 months were more likely to develop second malignancies following external beam radiotherapy than patients older than 12 months.219 Survivors of these second malignancies had a very high rate of developing subsequent tumors, particularly if they had been treated with radiation therapy.220 Fletcher and his colleagues221 have shown that second malignancies found in patients with bilateral retinoblastoma treated before the use of this modality were usually late-onset epithelial tumors, whereas second malignancies in patients following radiation therapy were usually earlier-onset sarcomas.
Plaque Radiotherapy
Plaque radiotherapy is a form of brachytherapy in which a radioactive implant is placed on the sclera over the base of a retinoblastoma with the intent of irradiating the tumor transclerally. The use of plaque radiotherapy is limited to tumors less than 16 mm in base and 8 mm in thickness. Effective treatment requires an average of 2 to 4 days of treatment time to deliver the total dose of 4,000 cGy to the apex of the tumor. Plaque radiotherapy can be used as either a primary treatment or a secondary treatment222,223,224,225 (Fig. 28.12). In 70% of cases, plaque radiotherapy is used as a secondary treatment to salvage a globe after failure of prior treatment, usually failed external beam radiotherapy or chemotherapy.222,223,224,225 In one series, solitary plaque radiotherapy was used in 91 cases of recurrent or residual retinoblastoma in which the only other option was enucleation.225 Tumor control and globe salvage was achieved in nearly 90% of these eyes.225
Figure 28.12 Plaque radiotherapy. A: Macular retinoblastoma before plaque radiotherapy. B: Regressed retinoblastoma after plaque radiotherapy.
Overall, there is nearly a 90% tumor control rate with one application of plaque radiotherapy.226 Carefully selected retinoblastomas, even juxtapapillary and macular tumors, can be successfully treated with plaque radiotherapy. The visual outcome for the patient varies with tumor size and location as well as associated radiation toxicity, which can include retinopathy or papillopathy. Positive visual outcomes have been reported in 62% of patients; the measured vision was 20/20 to 20/30 in more than one half of the cases.222 Radiation retinopathy and papillopathy become clinically manifest at approximately 18 months after irradiation, and these complications are more prominent in children who have been exposed to systemic chemotherapy. In an effort to avoid these problems with chemotherapy-treated patients, the tumor apex dose has been decreased to 3,500 cGy and radiation plaque therapy is delayed for at least 1 month after the child has discontinued chemotherapy. Innovations with custom design of plaques, especially those for small tumor recurrences, have also assisted in avoiding radiation retinopathy. Because of the use of focal, shielded radiation fields, plaque radiotherapy has not yet been found to be associated with induction of second cancers.
Laser Photocoagulation
Laser photocoagulation, using argon laser, diode laser, or xenon arc photocoagulation, can be used to treat small posterior retinoblastomas. Because the tumor size is important to the successful use of this treatment, tumors 4.5 mm or less in base and 2.5 mm or less in thickness with no evidence of vitreous seeds are usually selected.227,228 The treatment is directed to delimit the tumor and specifically coagulate all blood supply to the tumor. Two or three sessions at 1-month intervals are usually adequate to control most tumors. Use of the indirect ophthalmoscope laser photocoagulation system has greatly improved the facility of laser delivery.229 With laser treatment of properly selected cases of retinoblastoma, a 70% tumor control rate can be achieved. Recurrences are often treated with plaque radiotherapy. Complications of treatment include transient serous retinal detachment, visually significant retinal vascular occlusion, retinal traction, retinal hole, and preretinal fibrosis.
P.878

Cryotherapy
Cryotherapy is useful for managing equatorial and peripheral small retinoblastomas and is most successful if limited to tumors measuring 3.5 mm or less in diameter and 2.0 mm or less in thickness.230 Tumor destruction is usually achieved with one or two sessions of triple freeze-thaw cryotherapy at 1-month intervals. Cryotherapy will usually fail if there are overlying vitreous seeds. In these failed cases, plaque radiotherapy is usually used. Complications of cryotherapy include transient serous retinal detachment, retinal tear, localized preretinal fibrosis, and rhegmatogenous retinal detachment.231
Thermotherapy and Chemothermotherapy
Thermotherapy uses ultrasound, microwaves, or infrared radiation to deliver heat to the eye. The heat can be delivered to the whole eye with an attempt to spare the anterior segment,232 or the heat can be focused on one portion of the eye. The goal is to achieve a temperature of 42° to 60°C, a temperature that is below the coagulative threshold and, therefore, spares the retinal vessels of photocoagulation. The combination of heat and chemotherapy is termed chemothermotherapy, and the combination of heat and radiation is termed thermoradiotherapy. Heat has been found to have a synergistic effect with both chemotherapy and radiation therapy for the treatment of systemic and ocular cancers.233,234
The selection of the modality of thermotherapy or chemothermotherapy depends on many factors, including tumor size, location, laterality, status of the opposite eye, presence of subretinal fluid and seeds, presence of vitreous seeds, and prior or ongoing chemoreduction. Thermotherapy alone can often be used to effectively treat small retinoblastomas outside the retinal vascular arcades measuring 3 mm or less in size without vitreous or subretinal seeds. Thermotherapy alone without chemotherapy may be appropriate. The addition of other factors, such as larger tumors or seeds, often necessitates chemotherapy combined with thermotherapy for best tumor control.
When using thermotherapy alone, the goal is to heat the tumor to 45° to 60°C, which would leave a graywhite scar at the site. In general, small tumors require approximately 300 mW power for 10 minutes or less repeated for three times at 1-month intervals (Fig. 28.13). Tractional and vasoocclusive complications can occur within the retina because of the prolonged heating. When using chemothermotherapy, the goal is to heat the tumor to 42° to 45°C for 5 to 20 minutes, depending on the tumor size and location. Tumors up to 15 mm in base can be adequately treated with chemothermotherapy, especially if the patient is receiving three-agent chemoreduction. The result from chemothermotherapy is a light gray scar with less risk for tractional and retinal vascular problems than found with thermotherapy alone.233
Figure 28.13 Chemoreduction and chemothermotherapy. A: Large macular retinoblastoma overhanging the optic disc. B: After chemoreduction and chemothermotherapy, the tumor has regressed to a calcified scar.
There are several different chemothermotherapy protocols, each of which varies in chemotherapeutic agents and methods of delivery. Kaneko235 reported preliminary results using systemic and superselective ophthalmic artery injection of chemotherapy combined with thermotherapy. Murphree and Munier236 have used a specific protocol of intravenous carboplatin tightly coupled with thermotherapy. Shields and associates coupled thermotherapy within 4 hours of a chemoreduction regimen, thereby achieving the benefit of chemotherapy for both tumor reduction and consolidation. This method was more practical for those children with large or multiple tumors simultaneously on a chemoreduction protocol233,237,238,239 (Fig. 28.13). If a child is receiving chemoreduction, thermotherapy for tumor consolidation is generally initiated at cycle 2 or 3 of the chemoreduction protocol. Thermotherapy is repeated as necessary at each of the remaining chemoreduction cycles until six cycles are completed. Using this method for 188 retinoblastomas, complete tumor control in 86% of the tumors has been achieved.233 Success of this combined therapy is dependent on careful identification of suitable tumors. Smaller tumors without subretinal fluid or tumor seeds show the best response. In a recent study on the use of thermotherapy and chemothermotherapy for retinoblastoma, tumors less than 3 mm in base responded best with complete control and few complications.233 Tumors greater than 6 mm in base are at increased risk for recurrence of the main tumor or associated seeds and often require plaque radiotherapy.
The main complication of thermotherapy is focal iris atrophy related to heat effects on the pigmented iris tissue.233 In some instances, the lens develops a focal paraxial opacity. Chemothermotherapy is especially suited for small tumors adjacent to the fovea and optic nerve where radiation or laser photocoagulation would possibly induce more profound visual loss. This treatment modality is a time-consuming, tedious process that requires careful observations, recordings, judgments, and treatment adjustments in response to subtle tumor changes.
Intravenous Chemoreduction for Intraocular Retinoblastoma
Until recently, chemotherapy played only a minor role in the treatment of retinoblastoma. Chemotherapy was only used for patients in whom the disease had spread into the choroid, optic nerve, or orbit or to distant extraocular sites. In the past
P.879

few years, considerable experience has been gained in the use of chemotherapy for patients with intraocular retinoblastoma involving only the retina. The main objective of the ongoing clinical trials using chemotherapy in localized intraocular retinoblastoma has been to reduce the size of the tumors to an extent that would allow a variety of local surgical modalities such as laser photocoagulation, cryotherapy, or thermotherapy to control the residual disease. Successful management using chemotherapy in combination with local surgical methods can eliminate the use of external beam radiotherapy and therefore reduce significantly the risk of development of secondary malignancies and abnormalities of growth of orbital and facial bones associated with radiotherapy.219,240,241,242
The indications for chemoreduction in intraocular retinoblastoma are not clearly established. The goal of preserving vision in at least one eye while curing chil-dren with retinoblastoma makes patients with bilateral retinoblastoma the initial focal point for this form of therapy. A pilot study involving 40 eyes in 31 patients with bilateral disease who were treated with vincristine, teniposide, and carboplatin combined with cyclosporine resulted in a relapse-free rate of 89% in patients who were not previously treated.243 Relapse in this study was defined as tumor progression requiring either radiotherapy or enucleation. The median follow-up at the time of this report was 2 to 8 years. In another study of 20 patients who had 54 tumors in 31 eyes, a 2-month chemoreduction program with vincristine, carboplatin, and etoposide was followed by local treatment methods.244 Enucleation was avoided in all and external beam radiation therapy was necessary in only nine eyes because of diffuse vitreous seeds. Prior to therapy, the mean tumor base was 12 mm and the thickness 7 mm. Vitreous seeds were present in 14 eyes. After the 2-month chemotherapy regimen there was no evidence of residual viable tumor in 25 of 54 tumors. There was a mean decrease of 35% in tumor base and nearly 50% decrease in tumor thickness with resolution of subretinal fluid in 76% of cases. However, in the 14 eyes that had vitreal seeds, only 5 showed 90% to 100% calcification, indicating that different therapeutic modalities must be developed for the treatment of vitreous seeds. Careful follow-up of patients treated with chemoreduction is very important. Young patients with large tumors are at risk for recurrence of subretinal seeds. There is an increased risk for recurrence of retinal tumor and vitreous seeds in eyes with subretinal seeds at initial evaluation. These recurrences appear within 3 years of follow-up.245 In addition, new retinoblastomas develop at a mean interval of 5 months from initiation of chemoreduction and occur in 24% of patients by 5 years of follow-up. These new intraretinal tumors develop most commonly in patients who develop disease as young infants and who have a family history of retinoblastoma.246 Tumor size or location appears to be less important than the absence of subretinal tumor seeding and fluid and/or vitreous seeds for predicting successful globe salvage with chemoreduction accompanied by the use of local modalities247 (Table 28.3).
TABLE 28.3 COMPARISON OF REESE-ELLSWORTH (R-E) CLASSIFICATION AND THE PRACTICAL GROUPING (PG) SySTEM OF 158 EYES WITH RETINOBLASTOMA TREATED WITH CHEMOREDUCTION
R-E Classification No. (%) in Each
R-E Group
PG System No. (%) in Each
PG Group
I 9 (6) 1 38 (24)
II 26 (16) 2 11 (7)
III 16 (10) 3 75 (47)
IV 32 (20) 4 34 (22)
V 75 (47) 5a 0
aGroup 5 (high-risk retinoblastoma) currently is generally managed by primary enucleation and not chemoreduction.
Modified from Shields CL, Mashayekhi A, Demirci H, et al. Practical approach to management of retinoblastoma. Arch Ophthalmol 2004;122:729–735.
Chemotherapy for Possible Metastatic Disease
The treatment strategy for patients with intraocular retinoblastoma and extraretinal spread has not been conclusively studied. This includes patients with pathologic evidence of disease with extensive choroidal involvement, evidence of disease extending to the sclera, or disease extending beyond the lamina cribrosa but not yet involving the cut end of the optic nerve. There are conflicting reports about the significance of choroidal invasion. Kopelman et al.129 reported that choroidal invasion was not significantly associated with a fatal outcome. Messmer et al.185 reported that choroidal invasion was a low-risk factor and was not clinically significant when it was the only risk factor. The clinical significance increased considerably when choroidal invasion occurred in combination with other risk factors such as postlaminar optic nerve involvement, involvement of the optic nerve to the transection line, and late enucleation. Shields et al.188 reported that patients with choroidal invasion with any optic nerve invasion were at high risk for the development of metastases. However, in those patients with choroidal invasion alone, the risk for metastases was not significant. In this study, choroidal invasion was not classified as to extent of invasion, although others have reported that the degree of choroidal invasion assessed subjectively was an accurate predictor of survival.248 In the previous studies, some of the patients with choroid invasion with or without optic nerve invasion received radiotherapy and/or chemotherapy,
P.880

further clouding the interpretation of true metastatic potential of these risk factors. If significant choroidal invasion is present in the absence of optic nerve invasion, prophylactic adjuvant therapy may be considered. When extensive choroidal invasion is present in combination with optic nerve invasion beyond the lamina cribrosa, prophylactic adjuvant therapy is indicated.
Various chemotherapy regimens have been used for significant deep choroidal, optic nerve, ciliary body, or iris involvement. The combination of cyclophosphamide and doxorubicin has been used.249 Carboplatin, vincristine, and etoposide for 6 to 18 months has also been recommended.184,250 Recent retrospective studies suggest that prophylactic chemotherapy in patients with retinoblastoma with some of the previously mentioned high-risk features can minimize the risk of metastases.251,252,253 There are no randomized studies to recommend a particular regimen just as there are no randomized studies to identify specific criteria in patients with intraocular retinoblastoma with extraretinal extension who would benefit from chemotherapy.
Combined Radiotherapy and Chemotherapy
In three separate studies, patients presenting with Reese-Ellsworth eye group V retinoblastoma treated with radiotherapy alone have had 10%, 29%, and 66% of their eyes salvaged.209,214,254 In patients with bilateral retinoblastoma Reese-Ellsworth eye group V, Kingston et al.255 have shown that two cycles of chemotherapy in addition to external beam radiation therapy can preserve 70% of the eyes treated (14/20 eyes) with a median follow-up of 60 months. The study reporting a 66% salvage rate with radiotherapy alone and the study using chemotherapy in addition to radiotherapy were performed at the same institution; however, the patients in the latter study were reportedly more severely affected. Taken together, these data suggest that, for patients with retinoblastoma group V disease, the combination of chemotherapy with external beam radiotherapy may result in a salvage rate superior to that achieved with radiotherapy alone. In a recent study with a median follow-up of 13 months, chemotherapy alone, consisting of six cycles of carboplatin, vincristine, and etoposide, resulted in the salvage of only 50% of the eyes with groups IV and V disease without requiring external beam radiotherapy or enucleation.250 In the same study, chemotherapy eliminated the need for external beam radiotherapy or enucleation in all 39 eyes with groups I, II, and III disease (Table 28.4).
TABLE 28.4 PERCENTAGE OF GLOBES SALVAGED USING EXTERNAL BEAM RADIOTHERAPY ALONE, EXTERNAL BEAM RADIOTHERAPY AND SALVAGE TREATMENT, AND CHEMOREDUCTION AND FOCAL ADJUVANT TREATMENT
Reese-Ellsworth Group EBRT alone Ellsworth214
(1965–1972)
EBRT + salvage rx
Hungerford209 (1970–1985)
CRD + AT Shields203
(1994–1996)
I 91% 100% 100%
II 83% 84% 100%
III 82% 82% 100%
IV 62% 43% 100%
V 29% 66% 78%
EBRT, external beam radiotherapy; Rx, treatment; CRD, chemoreduction using vincristine, etoposide, and carboplatin; AT, adjuvant treatment (laser photocoagulation, cryotherapy, thermotherapy, chemothermotherapy, plaque radiotherapy, external beam radiotherapy).
CHEMOREDUCTION REGIMEN (6 CYCLES) FOR INTRAOCULAR RETINOBLASTOMA
Day Vincristine Etoposide Carboplatin
0 X X X
1   X  
Vincristine, 1.5 mg/m2, 0.05 mg/kg for children ≤36 months of age and maximum dose ≤2 mg; etoposide, 150 mg/m2, 5 mg/kg for children ≤36 months of age; carboplatin, 560 mg/m2, 18.6 mg/kg for children ≤36 months of age.
From the previous data it is clear that chemoreduction is an effective initial measure for selected children with intraocular retinoblastoma. Retinal tumor and seed recurrence remain a worrisome problem with chemoreduction.203 Seeds in the vitreous or subretinal space can recur in about 30% of eyes and enlarge to a visual and life-threatening state. Chemotherapy regimens used have included carboplatin, etoposide, and vincristine. Cyclosporine has been used with the previous regimen in an attempt to improve results by reversing multidrug resistance.256,257 Use of chemotherapy is not without concern, especially in patients with bilateral retinoblastoma, who have a higher incidence of second malignancies. The use of etoposide, an epipodophyllotoxin, has been associated with second malignancies in patients with leukemia and non-Hodgkin lymphoma. However, the schedule and the cumulative dose of etoposide used in most treatment regimens for retinoblastoma are different from those implicated in the incidence of second malignancies.258 Other side effects of concern are transient bone marrow suppression with a risk for infection.
P.881

Treatment of Systemic Retinoblastoma
Treatment of extraocular retinoblastoma requires a combined therapeutic approach using both chemotherapy and radiotherapy. Scleral involvement, orbital or bony involvement, involvement beyond the cut end of the optic nerve, metastatic disease involving brain or other sites, and trilateral retinoblastoma all require such an aggressive combined therapeutic approach. CNS involvement generally carries a poor prognosis.
Many different agents have been used to treat systemic retinoblastoma. One regimen used is the three-drug regimen including vincristine, carboplatin, and etoposide, similar to the chemoreduction regimen mentioned previously but with a much longer course of 6 to 18 months, depending on the clinical response.203,244 Others have found favorable results with similar chemotherapy regimens for extraocular retinoblastoma.259,260,261,262,263 White264 has reported on chemotherapy for retinoblastoma and recently advocated cyclophosphamide, etoposide, and vincristine as well as the support of peripheral stem cell rescue in multiple sequential courses for metastatic retinoblastoma. High-dose chemotherapy with autologous bone marrow or stem cell rescue appears to benefit patients with metastatic disease but without CNS involvement.265,266,267,268
Orbital Exenteration
Orbital exenteration is rarely used for retinoblastoma management in the United States because most retinoblastoma patients present with no evidence of extraocular invasion.269 Exenteration is most often used for orbital recurrence after the child has received a maximum acceptable dose of irradiation and chemotherapy. In other countries, patients may present with more advanced retinoblastoma, including orbital involvement. For these patients, exenteration, chemotherapy, and external beam radiotherapy are crucial for survival. Use of more advanced exenteration techniques, such as the eyelid-sparing exenteration, allows for rapid healing of the wound.270
Trilateral Retinoblastoma
Trilateral retinoblastoma is a term that describes the association of bilateral retinoblastoma and neuroblastic tumor in the pineal gland or other midline structures. A report from dePotter et al.271 revealed that this tumor occurs in children 4 years or younger. MRI or CT is essential to the diagnosis. The disease is fatal despite aggressive treatment with chemotherapy, radiation therapy, and gamma knife therapy. Longer survival has been correlated with earlier tumor diagnosis in asymptomatic patients. Trilateral retinoblastoma is a major cause of mortality in children within the first 5 years after diagnosis of bilateral retinoblastoma.272 No case of pinealoblastoma was observed in 147 children treated with initial chemoreduction who were followed up for 1 to 4 years. Although follow-up is limited, it is tempting to speculate that chemoreduction may reduce the risk for development of pinealoblastoma.273
HISTIOPATHOLOGIC CHANGES ASSOCIATED WITH TREATMENT
Enucleation of eyes from patients with retinoblastoma treated with different modalities that either fail to eradicate the tumor or result in major treatment complications allows the study of effects of therapy on the tumor and the ocular structures. Clinically, three types of regression patterns have been described in tumors that have undergone treatment: type 1 (cottage cheese), type 2 (fish flesh), or type 3 (combined).
In one study, five patients with sporadic bilateral retinoblastoma underwent planned enucleation of their functionally blind eye after two, three (in two patients), four, and six courses of primary chemotherapy with carboplatin, etoposide, cyclophosphamide, and vincristine. The eyes were examined histopathologically, using light microscopy and immunohistochemical analysis with proliferation markers. One patient had a type 1 (cottage cheese) regression and four patients had either a type 2 (fish flesh) or a type 3 (combined) regression pattern. Histopathologic examination revealed complete tumor necrosis with calcification (Fig. 28.14) in one patient with type 1 regression after three courses of chemotherapy and in one patient with type 3 regression after four courses of chemotherapy. The remaining three patients with type 2 or type 3 regression had histologic evidence of actively proliferating tumor cells after two, three, and six courses of chemotherapy. This report confirms histopathologically the
P.882

clinically described efficacy of primary chemotherapy in the treatment of retinoblastoma. The necessity for careful observation and the use of ancillary treatment whenever tumor regression (type 2 and 3 regression patterns) is not complete is, however, underscored.274 In another study, photoreceptor differentiation was observed in 17 of 42 enucleated eyes containing viable tumor following radiotherapy. Complications of external beam radiation include massive necrosis of the retina with associated hemorrhage (Fig. 28.14). Patients with retinoblastoma that have undergone chemotherapy and radiotherapy sometimes are left with tumors with large areas of fossilized cells and calcification with areas of photoreceptor differentiation similar to retinocytoma. Whether these cases represent chemotherapy or radiotherapy resistance of a focus of retinocytoma or differentiation induced by treatment has not been elucidated.275
Figure 28.14 A: Gross photograph of an autopsy eye in a patient with regressed retinoblastoma for more than 6 years after chemotherapy and radiotherapy. The patient died of complications of chemotherapy (cardiomyopathy). Notice the cottage cheese appearance (type 1 regression pattern) of the calcified (Ca++) tumor associated with proliferation of retinal pigment epithelium (*). B: Microscopic photograph of the same eye in (A) showing the calcified tumor (Ca++), proliferation of retinal pigment epithelium (*), and admixed glial tissue (hematoxylin and eosin stain, original magnification 20×). (See Color Figure 28.14.)
FUTURE DIRECTIONS
Subconjunctival Chemoreduction for Intraocular Retinoblastoma
To avoid the toxicity of systemic administration of chemotherapy, there is increasing interest in local delivery of these drugs to achieve chemoreduction for intraocular retinoblastoma. Studies in animal models show that carboplatin penetrates the sclera into the vitreous cavity, allowing for effective dosages within the eye with minimal toxicity.276,277,278 Local subconjunctival injection of carboplatin under protocol in humans as both a secondary treatment and a primary treatment has been used279 (Shields CL, Shields JA, unpublished observations). Within 3 to 4 weeks, tumor regression is usually noted, but the response may not be long term. Further investigations are necessary to determine the efficacy of local application of chemotherapy.
Chemotherapy to Prevent and Treat Metastases
The development of innovative modalities in the treatment of retinoblastoma has been hampered by the lack of suitable animal models of this uniquely human disease. Recently murine models of retinoblastoma using human xenografts in the vitreal space that mimic both metastatic and nonmetastatic disease280 as well as transgenic mouse models of retinoblastoma281 have been developed. These newer animal models will aid in the testing of new therapeutic approaches to retinoblastoma.
Gene Therapy
An alternative approach to local chemoreduction without the side effects of systemic chemotherapy is gene therapy. One form of gene therapy, suicide gene therapy, uses the delivery of the herpes simplex thymidine kinase gene by a replication-defective adenoviral vector injected directly into the tumor. Ganciclovir, a nucleoside analog, is then administered intravenously to the patients. The expressed thymidine kinase phosphorylates the ganciclovir within the tumor cells. The resulting nucleotide analog is a potent inhibitor of DNA synthesis and causes the death of the dividing tumor cells. Nondividing cells are left unaffected. The safety of this form of suicide gene therapy has been demonstrated in patients with brain tumors.282 Effective reduction of retinoblastoma tumors in a mouse model of the disease has also recently been demonstrated.283 A phase I clinical trial has been completed and this treatment has been shown to be safe and possibly effective for the treatment of children with retinoblastoma complicated by vitreous seeds.284
International Cooperative Studies
Because retinoblastoma is a very rare disease, clinical trials addressing basic diagnostic and therapeutic questions are very difficult to perform. Recently, the American College of Surgeons Oncology Group (ACOSOG) has formed a retinoblastoma study group to address therapeutic questions concerning the treatment of retinoblastoma.
ACKNOWLEDGMENTS
Support for this chapter was obtained from the National Institutes of Health (NCI and NEI), the Foundation for Research, and the Retina Research Foundation.
REFERENCES
1. Verhoeff F, Jackson E. Minutes of proceedings, 62nd annual meeting. Trans Am Ophthalmol Soc 1926;24:38.
2. Albert DM. Historic review of retinoblastoma. Ophthalmology 1987;94:654–662.
3. Albert DM. Wardrop Lecture, 1974. James Wardrop: a brief review of his life and contributions. Trans Ophthalmol Soc UK 1974;94:892–908.
4. Margo C, Hidayet A, Kopelman A, et al. Retinocytoma; a benign variant of retinoblastoma. Arch Ophthalmol 1983;101:1519–1531.
5. Gallie BL, Phillips RA, Ellsworth RM, et al. Significance of retinoma and phthisis bulbi for retinoblastoma. Ophthalmology 1982;89:1393–1399.
6. Bogenmann E, Lochrie MA, Simon MI. Cone cell-specific genes expressed in retinoblastoma. Science 1988;240:76–78.
7. Hurwitz RL, Bogenmann E, Font RL, et al. Expression of the functional cone phototransduction cascade in retinoblastoma. J Clin Invest 1990;85:1872–1878.
8. Donoso LA, Folberg R, Arbizo V. Retinal S antigen and retinoblastoma. A monoclonal antibody histopathologic study. Arch Ophthalmol 1985;103:855–857.
9. Nork TM, Millecchia LL, de Venecia GB, et al. Immunocytochemical features of retinoblastoma in an adult. Arch Ophthalmol 1996;114:1402–1406.
10. Nork TM, Schwartz TL, Doshi HM, et al. Retinoblastoma. Cell of origin. Arch Ophthalmol 1995;113:791–802.
11. Ohira A, Yamamoto M, Honda O, et al. Glial-, neuronal- and photoreceptor-specific cell markers in rosettes of retinoblastoma and retinal dysplasia. Curr Eye Res 1994;13:799–804.
12. Munier FL, Balmer A, Van Melle G, et al. Radial asymmetry in the topography of retinoblastoma. Clues to the cell of origin. Ophthalmic Genet 1994;15:101–106.
13. Yi YZ, Yang WZ, Zhen HL. [Retinoblastoma: cell origin and differentiation]. Chung Hua Yen Ko Tsa Chih 1994;30:214–217.
14. Rajagopalan S, Rodrigues MM, Wiggert B, et al. Retinoblastoma. Interphotoreceptor retinoid binding protein mRNA analysis by polymerase chain reaction. Ophthalmic Paediatr Genet 1993;14:117–125.
15. Chevez P, Font RL. Practical applications of some antibodies labelling the human retina. Histol Histopathol 1993;8:437–442.
16. Gonzalez-Fernandez F, Lopes MB, Garcia-Fernandez JM, et al. Expression of developmentally defined retinal phenotypes in the histogenesis of retinoblastoma. Am J Pathol 1992;141:363–375.
17. Kivela T. Parvalbumin, a horizontal cell-associated calcium-binding protein in retinoblastoma eyes. Invest Ophthalmol Vis Sci 1998;39:1044–1048.
18. McLean IW, Burnier M, Zimmerman L, Jakobiec F. Tumors of the Retina. In: McLean IW, Burnier MN, Zimmerman LE, FA J, eds. Atlas of tumor pathology. Tumors of the eye and ocular adnexa. Washington, DC: Armed Forced Institute of Pathology, 1994:100–135.
19. Petersen RA. Retinoblastoma. In: Albert D, Jakobiec F, eds. Principles and practice of ophthalmology: clinical practice. Philadelphia: Saunders, 1994:3279–3284.
20. Shields JA, Shields CL. Management and prognosis of retinoblastoma. Intraocular tumors. A text and atlas. Philadelphia: WB Saunders, 1992:377–392.
21. McLean IW. Retinoblastoma, retinocytomas, and pseudoretinoblastomas. In: Spencer WH, ed. Ophthalmic pathology. An atlas and textbook. Philadelphia: American Academy of Ophthalmology, WB Saunders, 1990:1332–1438.
P.883

22. Devesa S. The incidence of retinoblastoma. Am J Ophthalmol 1975;80:263–265.
23. Leal-Leal C, Flores-Rojo M, Medina-Sanson A, et al. A multicentre report from the Mexican Retinoblastoma Group. Br J Ophthalmol 2004;88:1074–1077.
24. Greene DM. Retinoblastoma. Diagnosis and management of malignant solid tumors in infants and children. Boston: Martinus Nijhoff, 1985:90.
25. Francois J, Matton M, deBie S, et al. Genesis and genetics of retinoblastoma. Ophthalmologica 1975;170:405.
26. Jensen RD, Miller RW. Retinoblastoma: epidemiologic characteristics. N Engl J Med 1971;285:307–311.
27. Schappert-Kimmiiser J, Hemmes GD, Nijiland R. The heredity of retinoblastoma. Ophthalmologica 1966;151:197–213.
28. Knudson AG Jr. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A 1971;68:820–823.
29. Bonaiti-Pellie C, Briard-Guillemot ML. Segregation analysis in hereditary retinoblastoma. Hum Genet 1981;57:411–419.
30. Hethcote HW, Knudson AG Jr. Model for the incidence of embryonal cancers: application to retinoblastoma. Proc Natl Acad Sci U S A 1978;75:2453–2457.
31. Lele KP, Penrose LS, Stallard HB. Chromosome deletion in a case of retinoblastoma. Ann Hum Genet 1963;27:171–174.
32. Chaum E, Ellsworth RM, Abramson DH, et al. Cytogenetic analysis of retinoblastoma: evidence for multifocal origin and in vivo gene amplification. Cytogenet Cell Genet 1984;38:82–91.
33. Turleau C, de Grouchy J, Chavin-Colin F, et al. Cytogenetic forms of retinoblastoma: their incidence in a survey of 66 patients. Cancer Genet Cytogenet 1985;16:321–334.
34. Squire J, Gallie BL, Phillips RA. A detailed analysis of chromosomal changes in heritable and non-heritable retinoblastoma. Hum Genet 1985;70:291–301.
35. Francke U. Retinoblastoma and chromosome 13. Cytogenet Cell Genet 1976;16:131–134.
36. Ward P, Packman S, Loughman W, et al. Location of the retinoblastoma susceptibility gene(s) and the human esterase D locus. J Med Genet 1984;21:92–95.
37. Harbour JW. Overview of RB gene mutations in patients with retinoblastoma. Implications for clinical genetic screening. Ophthalmology 1998;105:1442–1447.
38. Dryja TP, Cavenee W, White R, et al. Homozygosity of chromosome 13 in retinoblastoma. N Engl J Med 1984;310:550–553.
39. Godbout R, Dryja TP, Squire J, et al. Somatic inactivation of genes on chromosome 13 is a common event in retinoblastoma. Nature 1983;304:451–453.
40. Eng C, Li FP, Abramson DH, et al. Mortality from second tumors among long-term survivors of retinoblastoma [see comments]. J Natl Cancer Inst 1993;85:1121–1128.
41. Hansen MF, Koufos A, Gallie BL, et al. Osteosarcoma and retinoblastoma: a shared chromosomal mechanism revealing recessive predisposition. Proc Natl Acad Sci U S A 1985;82:6216–6220.
42. Dryja TP, Friend S, Weinberg RA. Genetic sequences that predispose to retinoblastoma and osteosarcoma. Symp Fundam Cancer Res 1986;39:115–119.
43. Huang HJ, Yee JK, Shew JY, et al. Suppression of the neoplastic phenotype by replacement of the RB gene in human cancer cells. Science 1988;242:1563–1566.
44. Takahashi R, Hashimoto T, Xu HJ, et al. The retinoblastoma gene functions as a growth and tumor suppressor in human bladder carcinoma cells. Proc Natl Acad Sci U S A 1991; 88:5257–5261.
45. Bookstein R, Lee EY, To H, et al. Human retinoblastoma susceptibility gene: genomic organization and analysis of heterozygous intragenic deletion mutants. Proc Natl Acad Sci U S A 1988;85:2210–2214.
46. Dryja TP, Rapaport JM, Epstein J, et al. Chromosome 13 homozygosity in osteosarcoma without retinoblastoma. Am J Hum Genet 1986;38:59–66.
47. Friend SH, Bernards R, Rogelj S, et al. A human DNA segment with properties of the gene that predisposes to retinoblastoma and osteosarcoma. Nature 1986;323:643–646.
48. Fung YK, Murphree AL, T’Ang A, et al. Structural evidence for the authenticity of the human retinoblastoma gene. Science 1987;236:1657–1661.
49. Dunn JM, Phillips RA, Zhu X, et al. Mutations in the RB1 gene and their effects on transcription. Mol Cell Biol 1989;9:4596–4604.
50. Yandell DW, Campbell TA, Dayton SH, et al. Oncogenic point mutations in the human retinoblastoma gene: their application to genetic counseling. N Engl J Med 1989;321:1689–1695.
51. Lohmann DR, Brandt B, Hopping W, et al. The spectrum of RB1 germ-line mutations in hereditary retinoblastoma. Am J Hum Genet 1996;58:940–949.
52. Yilmaz S, Horsthemke B, Lohmann DR. Twelve novel RB1 gene mutations in patients with hereditary retinoblastoma. Mutations in brief no. 206. Online. Hum Mutat 1998; 12:434.
53. Blanquet V, Turleau C, Gross-Morand MS, et al. Spectrum of germline mutations in the RB1 gene: a study of 232 patients with hereditary and non hereditary retinoblastoma. Hum Mol Genet 1995;4:383–388.
54. Cowell JK, Jaju R, Kempski H. Isolation and characterisation of a panel of cosmids which allows unequivocal identification of chromosome deletions involving the RB1 gene using fluorescence in situ hybridisation. J Med Genet 1994;31:334–337.
55. Szijan I, Lohmann DR, Parma DL, et al. Identification of RB1 germline mutations in Argentinian families with sporadic bilateral retinoblastoma. J Med Genet 1995;32:475–479.
56. Dryja TP, Mukai S, Petersen R, et al. Parental origin of mutations of the retinoblastoma gene. Nature 1989;339:556–558.
57. Zhu XP, Dunn JM, Phillips RA, et al. Preferential germline mutation of the paternal allele in retinoblastoma. Nature 1989;340:312–313.
58. Toguchida J, Ishizaki K, Sasaki MS, et al. Preferential mutation of paternally derived RB gene as the initial event in sporadic osteosarcoma. Nature 1989;338:156–158.
59. Harbour JW, Lai S-L, Whang-Peng J, et al. Abnormalities in structure and expression of the human retinoblastoma gene in SCLC. Science 1988;241:353–357.
60. T’Ang A, Varley JM, Chakraborty S, et al. Structural rearrangement of the retinoblastoma gene in human breast carcinoma. Science 1988;242:263–266.
61. Bookstein R, Lee EY, Peccei A, et al. Human retinoblastoma gene: long-range mapping and analysis of its deletion in a breast cancer cell line. Mol Cell Biol 1989;9:1628–1634.
62. Seizinger BR, Klinger HP, Junien C, et al. Report of the committee on chromosome and gene loss in human neoplasia. Cytogenet Cell Genet 1991;58:1080–1096.
63. Lee W-H, Bookstein R, Hong F, et al. Human retinoblastoma susceptibility gene: cloning, identification, and sequence. Science 1987;235:1394–1399.
64. Lee WH, Shew JY, Hong FD, et al. The retinoblastoma susceptibility gene encodes a nuclear phosphoprotein associated with DNA binding activity. Nature 1987;329:642–645.
65. Buchkovich K, Duffy LA, Harlow E. The retinoblastoma protein is phosphorylated during specific phases of the cell cycle. Cell 1989;58:1097–1105.
66. Magnaghi-Jaulin L, Groisman R, Naguibneva I, et al. Retinoblastoma protein represses transcription by recruiting a histone deacetylase [see comments]. Nature 1998;391:601–605.
67. Brehm A, Miska EA, McCance DJ, et al. Retinoblastoma protein recruits histone deacetylase to repress transcription [see comments]. Nature 1998;391:597–601.
68. Whyte P, Buchkovich KJ, Horowitz JM, et al. Association between an oncogene and an anti-oncogene: the adenovirus E1A proteins bind to the retinoblastoma gene product. Nature 1988;334:124–129.
69. DeCaprio JA, Ludlow JW, Figge J, et al. SV40 large tumor antigen forms a specific complex with the product of the retinoblastoma susceptibility gene. Cell 1988;54:275–283.
70. Dyson N, Howley PM, Münger K, et al. The human papilloma virus-16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science 1989;243:934–937.
71. Vogel F. Genetics of retinoblastoma. Human Genetics 1979;52:1–54.
72. Bremner R, Du DC, Connolly-Wilson MJ, et al. Deletion of RB exons 24 and 25 causes low-penetrance retinoblastoma. Am J Hum Genet 1997;61:556–570.
73. Sippel KC, Fraioli RE, Smith GD, et al. Frequency of somatic and germ-line mosaicism in retinoblastoma: implications for genetic counseling. Am J Hum Genet 1998;62:610–619.
74. Zimmerman L. Retinoblastoma and retinocytoma. In: Spencer, WH, ed., Ophthalmic pathology. An atlas and textbook. American Academy of Ophthalmology. Philadelphia: WB Saunders, 1985:1292–1351.
75. Margo CE, Zimmerman LE. Retinoblastoma: the accuracy of clinical diagnosis in children treated by enucleation. J Pediatr Ophthalmol Strabismus 1983;20:227–229.
76. Shields CL, Shields JA, Shields MB, et al. Prevalence and mechanisms of secondary intraocular pressure elevation in eyes with intraocular tumors. Ophthalmology 1987;94:839–846.
77. Shields JA, Shields CL, Parsons HM. Differential diagnosis of retinoblastoma. Retina 1991;11:232–243.
78. Shields JA, Augsburger JJ. Current approaches to the diagnosis and management of retinoblastoma. Surv Ophthalmol 1981;25:347–372.
79. Yoshizumi MO, Thomas JV, Smith TR. Glaucoma-inducing mechanisms in eyes with retinoblastoma. Arch Ophthalmol 1978;96:105–110.
80. Arrigg PG, Hedges TR III, Char DH. Computed tomography in the diagnosis of retinoblastoma. Br J Ophthalmol 1983;67:588–591.
81. Smith EV, Gragoudas ES, Kolodny NH, et al. Magnetic resonance imaging: an emerging technique for the diagnosis of ocular disorders. Int Ophthalmol 1990;14:119–124.
82. Chang MM, McLean IW, Merritt JC. Coats’ disease: a study of 62 histologically confirmed cases. J Pediatr Ophthalmol Strabismus 1984;21:163–168.
83. Morgan KS, McLean IW. Retinoblastoma and persistent hyperplastic vitreous occurring in the same patient. Ophthalmology 1981;88:1087–1089.
84. Gassler N, Lommatzsch PK. [Clinicopathologic study of 817 enucleations]. Klin Monatsbl Augenheilkd 1995;207:295–301.
85. Wieckowska A, Napierala A, Pytlarz E, et al. [Persistent hyperplastic primary vitreous—diagnosis and differentiation]. Klin Oczna 1995;97:234–238.
86. Steidl SM, Hirose T, Sang D, et al. Difficulties in excluding the diagnosis of retinoblastoma in cases of advanced Coats’ disease: a clinicopathologic report. Ophthalmologica 1996;210:336–340.
87. Ells A, Clarke WN, Noel LP. Pseudohypopyon in acute myelogeneous leukemia. J Pediatr Ophthalmol Strabismus 1995;32:123–124.
88. Riss JM, Girard NJ, Proust H, et al. Diffuse choroidal hemangioma: report of a clinicopathological study in a 4-year-old boy. Ophthalmologica 1995;209:284–288.
89. Hanssens M, Meire F. Pseudoglioma: a clinico-pathological report [clinical conference]. Bull Soc Belge Ophtalmol 1995;255:99–105.
90. Smirniotopoulos JG, Bargallo N, Mafee MF. Differential diagnosis of leukokoria: radiologic-pathologic correlation. Radiographics 1994;14:1059–1079.
91. Tajima Y, Nakajima T, Sugano I, et al. Cytodiagnostic clues to primary retinoblastoma based on cytologic and histologic correlates of 39 enucleated eyes. Acta Cytol 1994;38:151–157.
92. Caruso J, Miller KB, Pietrantonio JJ. Combined hamartoma of the retina and retinal pigment epithelium. Optom Vis Sci 1993;70:860–862.
93. Scott MH, Richard JM. Retinoblastoma in the state of Oklahoma: a clinicopathologic review. J Okla State Med Assoc 1993;86:111–118.
94. Minoda K, Hirose Y, Sugano I, et al. Occurrence of sequential intraocular tumors: malignant medulloepithelioma subsequent to retinoblastoma. Jpn J Ophthalmol 1993;37:293–300.
95. Sharma A, Ram J, Gupta A. Solitary retinal astrocytoma. Acta Ophthalmol (Copenh) 1991;69:113–116.
96. Hausmann N, Stefani FH. Medulloepithelioma of the ciliary body. Acta Ophthalmol (Copenh) 1991;69:398–401.
97. Kuker W, Ramaekers V. Persistent hyperplastic primary vitreous: MRI. Neuroradiology 1999;41:520–522.
98. Kaste SC, Jenkins JJ III, Meyer D, et al. Persistent hyperplastic primary vitreous of the eye: imaging findings with pathologic correlation. AJR Am J Roentgenol 1994;162:437–440.
99. Liang JC, Augsburger JJ, Shields JA. Diffuse infiltrating retinoblastoma associated with persistent primary vitreous. J Pediatr Ophthalmol Strabismus 1985;22:31–33.
P.884

100. Haddad R, Font RL, Reeser F. Persistent hyperplastic primary vitreous. A clinicopathologic study of 62 cases and review of the literature. Surv Ophthalmol 1978;23:123–134.
101. Irvine AR, Albert DM, Sang DN. Retinal neoplasia and dysplasia. II. Retinoblastoma occurring with persistence and hyperplasia of the primary vitreous. Invest Ophthalmol Vis Sci 1977;16:403–407.
102. Gunalp I, Gunduz K, Arslan Y. Retinoblastoma in Turkey—treatment and prognosis. Jpn J Ophthalmol 1996;40:95–102.
103. Stewart J, Halliwell T, Gupta RK. Cytodiagnosis of Coats’ disease from an ocular aspirate. A case report. Acta Cytol 1993;37:717–720.
104. Kremer I, Nissenkorn I, Ben-Sira I. Cytologic and biochemical examination of the subretinal fluid in diagnosis of Coats’ disease. Acta Ophthalmol (Copenh) 1989;67:342–346.
105. Haik BG, Koizumi J, Smith ME, et al. Fresh preparation of subretinal fluid aspirations in Coats’ disease. Am J Ophthalmol 1985;100:327–328.
106. Manschot WA, de Bruijn WC. Coats’s disease: definition and pathogenesis. Br J Ophthalmol 1967;51:145–157.
107. Sherman JL, McLean IW, Brallier DR. Coats’ disease: CT-pathologic correlation in two cases. Radiology 1983;146:77–78.
108. Katz NN, Margo CE, Dorwart RH. Computed tomography with histopathologic correlation in children with leukokoria. J Pediatr Ophthalmol Strabismus 1984;21:50–56.
109. Potter PD, Shields CL, Shields JA, et al. The role of magnetic resonance imaging in children with intraocular tumors and simulating lesions. Ophthalmology 1996;103:1774–1783.
110. Edward DP, Mafee MF, Garcia-Valenzuela E, et al. Coats’ disease and persistent hyperplastic primary vitreous. Role of MR imaging and CT. Radiol Clin North Am 1998;36:1119–1131, x.
111. Wycliffe ND, Mafee MF. Magnetic resonance imaging in ocular pathology. Top Magn Reson Imaging 1999;10:384–400.
112. Roth AM. Retinoblastoma seen after surgery for traumatic cataract. Ann Ophthalmol 1978;10:1561–1564.
113. Shields CL, Honavar S, Shields JA, et al. Vitrectomy in eyes with unsuspected retinoblastoma. Ophthalmology 2000;107:2250–2255.
114. Char DH, Miller TR. Fine needle biopsy in retinoblastoma. Am J Ophthalmol 1984;97:686–690.
115. Akhtar M, Ali MA, Sabbah R, et al. Fine-needle aspiration biopsy diagnosis of round cell malignant tumors of childhood. A combined light and electron microscopic approach. Cancer 1985;55:1805–1817.
116. Alio J, Ludena M, Millan A, et al. Ultrastructural study of a retinoma by intraocular fine-needle aspiration biopsy. Ophthalmologica 1988;196:192–199.
117. Akhtar M, Ali MA, Sabbah R, et al. Aspiration cytology of retinoblastoma: light and electron microscopic correlations. Diagn Cytopathol 1988;4:306–311.
118. Das DK, Das J, Chachra KL, et al. Diagnosis of retinoblastoma by fine-needle aspiration and aqueous cytology. Diagn Cytopathol 1989;5:203–206.
119. Shields JA, Shields CL, Ehya H, et al. Fine-needle aspiration biopsy of suspected intraocular tumors. The 1992 Urwick Lecture. Ophthalmology 1993;100:1677–1684.
120. Robertson DM. Fine-needle biopsy and retinoblastoma [letter; comment]. Ophthalmology 1997;104:567–568.
121. Decaussin M, Boran MD, Salle M, et al. Cytological aspiration of intraocular retinoblastoma in an 11-year-old boy. Diagn Cytopathol 1998;19:190–193.
122. Karcioglu ZA. Fine needle aspiration biopsy (FNAB) for retinoblastoma. Retina 2002;22: 707–710.
123. O’hara BJ, Ehya H, Shields JA, et al. Fine needle aspiration biopsy in pediatric ophthalmic tumors and pseudotumors. Acta Cytol 1993;37:125–130.
124. Augsburger JJ, Shields JA, Folberg R, et al. Fine needle aspiration biopsy in the diagnosis of intraocular cancer. Cytologic-histologic correlations. Ophthalmology 1985;92:39–49.
125. Spencer WH. Optic nerve extension of intraocular neoplasms. Am J Ophthalmol 1975;80:465–471.
126. Karcioglu ZA, al Mesfer SA, Abboud E, et al. Workup for metastatic retinoblastoma. A review of 261 patients. Ophthalmology 1997;104:307–312.
127. Tosi P, Cintorino M, Toti P, et al. Histopathological evaluation for the prognosis of retinoblastoma. Ophthalmic Paediatr Genet 1989;10:173–177.
128. Croxatto JO, Fernandez Meijide R, Malbran ES. Retinoblastoma masquerading as ocular inflammation. Ophthalmologica 1983;186:48–53.
129. Kopelman JE, McLean IW, Rosenby SH. Multivariate analysis of risk factors for metastasis in retinoblastoma treated by enucleation. Ophthalmology 1987;94:371–377.
130. Donaldson SS, Smith LM. Retinoblastoma: biology, presentation, and current management. Oncology (Huntingt) 1989;3:45–51.
131. Grossniklaus HE, Dhaliwal RS, Martin DF. Diffuse anterior retinoblastoma. Retina 1998;18:238–241.
132. Moll AC, Koten JW, Lindenmayer DA, et al. Three histopathological types of retinoblastoma and their relation to heredity and age of enucleation. J Med Genet 1996;33:923–927.
133. Zilelioglu G, Gunduz K. Ultrasonic findings in intraocular retinoblastoma and correlation with histopathologic diagnosis. Int Ophthalmol 1995;19:71–75.
134. Nemeth J, Szabo A, Vegh M. Unusual echographic form of retinoblastoma. Acta Ophthalmol Suppl 1992;204:107–109.
135. Bhatnagar R, Vine AK. Diffuse infiltrating retinoblastoma. Ophthalmology 1991;98:1657–1661.
136. Mansour AM, Greenwald MJ, O’Grady R. Diffuse infiltrating retinoblastoma. J Pediatr Ophthalmol Strabismus 1989;26:152–154.
137. Girard B, Le Hoang P, D’Hermies F, et al. [Diffuse infiltrating retinoblastoma]. J Fr Ophtalmol 1989;12:369–381.
138. Shields JA, Shields CL, Eagle RC, et al. Spontaneous pseudohypopyon secondary to diffuse infiltrating retinoblastoma. Arch Ophthalmol 1988;106:1301–1302.
139. Nicholson DH, Norton EW. Diffuse infiltrating retinoblastoma. Trans Am Ophthalmol Soc 1980;78:265–289.
140. Morgan G. Diffuse infiltrating retinoblastoma. Br J Ophthalmol 1971;55:600–606.
141. Lindley J, Smith S. Histology and spontaneous regression of retinoblastoma. Trans Ophthalmol Soc U K 1974;94:953–967.
142. Andersen SR, Jensen OA. Retinoblastoma with necrosis of central retinal artery and vein and partial spontaneous regression. Acta Ophthalmol 1974;52:183–193.
143. Galimova RZ, Zuikova TP, Buriakova ZA. [Clinico-morphological features of retinoblastoma with spontaneous regression]. Vestn Oftalmol 1990;106:56–59.
144. Greger V, Passarge E, Hopping W, et al. Epigenetic changes may contribute to the formation and spontaneous regression of retinoblastoma. Hum Genet 1989;83:155–158.
145. Krasnovid TA. [Spontaneous regression of bilateral retinoblastoma]. Oftalmol Zh 1987;4: 248–249.
146. Assaf AA, Phillips CI. Spontaneous regression of unilateral retinoblastoma in a father of three sons with bilateral retinoblastoma. Ophthalmic Paediatr Genet 1985;6:179–182.
147. Gangwar DN, Jain IS, Gupta A, et al. Bilateral spontaneous regression of retinoblastoma with dominant transmission. Ann Ophthalmol 1982;14:479–480.
148. Brodwall J. Spontaneous regression of a retinoblastoma. A case report. Acta Ophthalmol (Copenh) 1981;59:430–434.
149. Khodadoust AA, Roozitalab HM, Smith RE, et al. Spontaneous regression of retinoblastoma. Surv Ophthalmol 1977;21:467–478.
150. Nehen JH. Spontaneous regression of retinoblastoma. Acta Ophthalmol (Copenh) 1975;53:647–651.
151. Pearce WG, Gillan JG. Bilateral spontaneous regression of retinoblastoma. Can J Ophthalmol 1972;7:234–239.
152. Karsgaard AT. Spontaneous regression of retinoblastoma. A report of two cases. Can J Ophthalmol 1971;6:218–222.
153. Boniuk M, Bishop DW. Oligodendroglioma of the retina. Surv Ophthalmol 1969;13:284–289.
154. Mullaney PB, Karcioglu ZA, Huaman AM, et al. Retinoblastoma associated orbital cellulitis. Br J Ophthalmol 1998;82:517–521.
155. Salazar-Flores M, Ambrosius-Diener K. [Retinoblastoma. Anato-mical study of 406 cases]. Bol Med Hosp Infant Mex 1986;43:106–112.
156. Shuangshoti S, Chaiwun B, Kasantikul V. A study of 39 retinoblastomas with particular reference to morphology, cellular differentiation and tumour origin. Histopathology 1989;15:113–124.
157. Lamping KA, Albert DM, Snyder C, et al. The Harrower collection and its place in the history of ophthalmic pathology. Surv Ophthalmol 1983;27:374–380.
158. Sang DN, Albert DM. Retinoblastoma: clinical and histopathologic features. Hum Pathol 1982;13:133–147.
159. Bierring F, Egeberg J, Jensen OA. A contribution to the ultrastructural study of retinoblastomas. Acta Ophthalmol 1967;45:424–428.
160. Datta BN. DNA coating of blood vessels in retinoblastomas. Am J Clin Pathol 1974;62:94–96.
161. Radnot M. Scanning electron microscopy of retinoblastoma. J Pediatr Ophthalmol Strabismus 1978;15:36–39.
162. Ts’o MO, Fine BS, Zimmerman LE. The Flexner-Wintersteiner rosettes in retinoblastoma. Arch Pathol 1969;88:664–671.
163. Vrabec T, Arbizo V, Adamus G, et al. Rod cell-specific antigens in retinoblastoma. Arch Ophthalmol 1989;107:1061–1063.
164. Abramson DH, Greenfield DS, Ellsworth RM, et al. Neuron-specific enolase and retinoblastoma. Clinicopathologic correlations. Retina 1989;9:148–152.
165. Bardenstein DS, Rodrigues MM, Alroy J, et al. Lectin binding in retinoblastoma. Curr Eye Res 1987;6:1141–1150.
166. Kivela T, Tarkkanen A. S-100 protein in retinoblastoma revisited. An immunohistochemical study. Acta Ophthalmol (Copenh) 1986;64:664–673.
167. Rodrigues MM, Wilson ME, Wiggert B, et al. Retinoblastoma. A clinical, immunohistochemical, and electron microscopic case report. Ophthalmology 1986;93:1010–1015.
168. Donoso LA, Felberg NT, Augsburger JJ, et al. Retinal S-antigen and retinoblastoma: a monoclonal antibody and flow cytometric study. Invest Ophthalmol Vis Sci 1985;26:568–571.
169. Zhang J, Gray J, Wu L, et al. Rb regulates proliferation and rod photoreceptor development in the mouse retina. Nat Genet 2004;36:351–360.
170. Eagle RC, Jr., Shields JA, Donoso L, et al. Malignant transformation of spontaneously regressed retinoblastoma, retinoma/retinocytoma variant. Ophthalmology 1989;96:1389–1395.
171. Spraul CW, Lim JI, Lambert SR, et al. Retinoblastoma recurrence after iodine 125 plaque application. Retina 1996;16:135–138.
172. Ts’o MO, Fine BS, Zimmerman LE. The nature of retinoblastoma. II. Photoreceptor differentiation: an electron microscopic study. Am J Ophthalmol 1970;69:350–359.
173. Ts’o MO, Zimmerman LE, Fine BS. The nature of retinoblastoma. I. Photoreceptor differentiation: a clinical and histopathologic study. Am J Ophthalmol 1970;69:339–349.
174. Ts’o MO, Zimmerman LE, Fine BS, et al. A cause of radioresistance in retinoblastoma: photoreceptor differentiation. Trans Am Acad Ophthalmol Otolaryngol 1970;74:959–969.
175. Singh AD, Santos CM, Shields CL, et al. Observations on 17 patients with retinocytoma. Arch Ophthalmol 2000;118:199–205.
176. Benhamou E, Borges J, Tso MO. Magnetic resonance imaging in retinoblastoma and retinocytoma: a case report. J Pediatr Ophthalmol Strabismus 1989;26:276–280.
177. 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;77:1206–1213.
178. Kivela T. Trilateral retinoblastoma: a meta-analysis of hereditary retinoblastoma associated with primary ectopic intracranial retinoblastoma [see comments]. J Clin Oncol 1999;17:1829–1837.
179. Reese AB, Ellsworth RM. Management of retinoblastoma. Ann N Y Acad Sci 1964;114:958–962.
180. Wintersteiner H. Das Neuroepithelioma Retinae. Leipzig und Wien: Franz Doeticke, 1897.
181. Magramm I, Abramson DH, Ellsworth RM. Optic nerve involvement in retinoblastoma. Ophthalmology 1989;96:217–222.
P.885

182. Mustafa MM, Jamshed A, Khafaga Y, et al. Adjuvant chemotherapy with vincristine, doxorubicin, and cyclophosphamide in the treatment of postenucleation high risk retinoblastoma. J Pediatr Hematol Oncol 1999;21:364–369.
183. Chantada GL, de Davila MT, Fandino A, et al. Retinoblastoma with low risk for extraocular relapse. Ophthalmic Genet 1999;20:133–140.
184. Gunduz K, Shields CL, Shields JA, et al. The outcome of chemoreduction treatment in patients with Reese-Ellsworth group V retinoblastoma. Arch Ophthalmol 1998;116:1613–1617.
185. Messmer EP, Heinrich T, Hopping W, et al. Risk factors for metastases in patients with retinoblastoma. Ophthalmology 1991;98:136–141.
186. Messmer EP, Fritze H, Mohr C, et al. Long-term treatment effects in patients with bilateral retinoblastoma: ocular and mid-facial findings. Graefes Arch Clin Exp Ophthalmol 1991;229:309–314.
187. Margo C, Hidayat AA, Marshall CF, et al. Cryotherapy and photocoagulation in the management of retinoblastoma: treatment failure and unusual complication. Ophthalmic Surg 1983;14:336–342.
188. Shields CL, Shields JA, Baez KA, et al. Choroidal invasion of retinoblastoma: metastatic potential and clinical risk factors [see comments]. Br J Ophthalmol 1993;77:544–548.
189. Shields CL, Shields JA, Baez K, et al. Optic nerve invasion of retinoblastoma: metastatic potential and clinical risk factors. Cancer 1994;73:692–698.
190. Mohney BG, Robertson DM. Ancillary testing for metastasis in patients with newly diagnosed retinoblastoma. Am J Ophthalmol 1994;118:707–711.
191. Survival rate and risk factors for patients with retinoblastoma in Japan. The Committee for the National Registry of Retinoblastoma. Jpn J Ophthalmol 1992;36:121–131.
192. Erwenne CM, Franco EL. Age and lateness of referral as determinants of extra-ocular retinoblastoma. Ophthalmic Paediatr Genet 1989;10:179–184.
193. Rubin CM, Robison LL, Camerson JD, et al. Intraocular retinoblastoma Group V: an analysis of prognostic factors. J Clin Oncol 1985;3:680–685.
194. MacKay CJ, Abramson DH, Ellsworth RM. Metastatic patterns of retinoblastoma. Arch Ophthalmol 1984;102:391–396.
195. Rootman J, Ellsworth RM, Hofbauer J, et al. Orbital extension of retinoblastoma: a clinicopathological study. Can J Ophthalmol 1978;13:72–80.
196. de Buen S, Gonzalez-Almaraz G, Cruz-Perez R. [Retinoblastoma. Considerations on its biological behavior]. Gac Med Mex 1974;108:177–186.
197. Stannard C, Lipper S, Sealy R, et al. Retinoblastoma: correlation of invasion of the optic nerve and choroid with prognosis and metastases. Br J Ophthalmol 1979;63:560–570.
198. Shields CL, Shields JA. Recent developments in the management of retinoblastoma. J Pediatr Ophthalmol Strabismus 1999;36:8–18.
199. Augsburger JJ, Oehlschlager U, Manzitti JE. Multinational clinical and pathologic registry of retinoblastoma. Retinoblastoma International Collaborative Study report 2. Graefes Arch Clin Exp Ophthalmol 1995;233:469–475.
200. Shields JA, Shields CL. Current management of retinoblastoma. Mayo Clin Proc 1994;69:50–56.
201. Shields JA. Misconceptions and techniques in the management of retinoblastoma. The 1992 Paul Henkind Memorial Lecture. Retina 1992;12:320–330.
202. Dudgeon J. Retinoblastoma—trends in conservative management [editorial; comment]. Br J Ophthalmol 1995;79:104.
203. Shields CL, Shields JA, Needle M, et al. Combined chemoreduction and adjuvant treatment for intraocular retinoblastoma [see comments]. Ophthalmology 1997;104:2101–2111.
204. Shields JA, Shields CL, Sivalingam V. Decreasing frequency of enucleation in patients with retinoblastoma. Am J Ophthalmol 1989;108:185–188.
205. Shields JA, Shields CL, Eagle RC, et al. Calcified intraocular abscess simulating retinoblastoma [letter]. Am J Ophthalmol 1992;114:227–229.
206. Shields JA, Shields CL, de Potter P. Enucleation technique for children with retinoblastoma. J Pediatr Ophthalmol Strabismus 1992;29:213–215.
207. Karcioglu ZA, al Mesfer SA, Mullaney PB. Porous polyethylene orbital implant in patients with retinoblastoma. Ophthalmology 1998;105:1311–1316.
208. Shields CL, Shields JA, de Potter P, et al. Lack of complications of the hydroxyapatite orbital implant in 250 consecutive cases. Trans Am Ophthalmol Soc 1993;91:177–189.
209. Hungerford JL, Toma NM, Plowman PN, et al. External beam radiotherapy for retinoblastoma: I. Whole eye technique [see comments]. Br J Ophthalmol 1995;79:109–111.
210. Toma NM, Hungerford JL, Plowman PN, et al. External beam radiotherapy for retinoblastoma: II. Lens sparing technique [see comments]. Br J Ophthalmol 1995;79:112–117.
211. Singh AD, Garway-Heath D, Love S, et al. Relationship of regression pattern to recurrence in retinoblastoma. Br J Ophthalmol 1993;77:12–16.
212. Abramson DH, Servodidio CA, De Lillo AR, et al. Recurrence of unilateral retinoblastoma following radiation therapy. Ophthalmic Genet 1994;15:107–113.
213. Fontanesi J, Pratt CB, Hustu HO, et al. Use of irradiation for therapy of retinoblastoma in children more than 1 year old: the St. Jude Children’s Research Hospital experience and review of literature. Med Pediatr Oncol 1995;24:321–326.
214. Ellsworth RM. Retinoblastoma. Mod Probl Ophthalmol 1977;18:94–100.
215. Plowman PN, Kingston JE, Hungerford JL. Prophylactic retinal radiotherapy has an exceptional place in the management of familial retinoblastoma [see comments]. Br J Cancer 1993;68:743–745.
216. Brooks HL Jr, Meyer D, Shields JA, et al. Removal of radiation-induced cataracts in patients treated for retinoblastoma. Arch Ophthalmol 1990;108:1701–1708.
217. Weiss AH, Karr DJ, Kalina RE, et al. Visual outcomes of macular retinoblastoma after external beam radiation therapy. Ophthalmology 1994;101:1244–1249.
218. Roarty JD, McLean IW, Zimmerman LE. Incidence of second neoplasms in patients with bilateral retinoblastoma. Ophthalmology 1988;95:1583.
219. Abramson DH, Frank CM. Second nonocular tumors in survivors of bilateral retinoblastoma: a possible age effect on radiation-related risk [see comments]. Ophthalmology 1998;105:573–579.
220. Abramson DH, Melson MR, Dunkel IJ, et al. Third (fourth and fifth) nonocular tumors in survivors of retinoblastoma. Ophthalmology 2001;108:1868–1876.
221. Fletcher O, Easton D, Anderson K, et al. Lifetime risks of common cancers among retinoblastoma survivors. J Natl Cancer Inst 2004;96:357–363.
222. Shields CL, Shields JA, de Potter P, et al. Plaque radiotherapy in the management of retinoblastoma. Use as a primary and secondary treatment [see comments]. Ophthalmology 1993;100:216–224.
223. Hernandez JC, Brady LW, Shields CL, et al. Conservative treatment of retinoblastoma. The use of plaque brachytherapy. Am J Clin Oncol 1993;16:397–401.
224. Desjardins L, Levy C, Labib A, et al. An experience of the use of radioactive plaques after failure of external beam radiation in the treatment of retinoblastoma. Ophthalmic Paediatr Genet 1993;14:39–42.
225. Shields JA, Shields CL, de Potter P, et al. Plaque radiotherapy for residual or recurrent retinoblastoma in 91 cases. J Pediatr Ophthalmol Strabismus 1994;31:242–245.
226. Shields CL, Shields JA, Minelli S, et al. Regression of retinoblastoma after plaque radiotherapy [see comments]. Am J Ophthalmol 1993;115:181–187.
227. Shields JA. The expanding role of laser photocoagulation for intraocular tumors. 1993 H. Christian Zweng Memorial Lecture. Retina 1994;14:310–322.
228. Shields JA, Shields CL, Parsons H, et al. The role of photocoagulation in the management of retinoblastoma. Arch Ophthalmol 1990;108:205–208.
229. Shields CL, Shields JA, Kiratli H, et al. Treatment of retinoblastoma with indirect ophthalmoscope laser photocoagulation. J Pediatr Ophthalmol Strabismus 1995;32:317–322.
230. Shields JA, Parsons H, Shields CL, et al. The role of cryotherapy in the management of retinoblastoma. Am J Ophthalmol 1989;106:260–264.
231. Baumal CR, Shields CL, Shields JA, et al. Surgical repair of rhegmatogenous retinal detachment after treatment for retinoblastoma. Ophthalmology 1998;105:2134–2139.
232. Lagendijk JJ. A microwave heating technique for the hyperthermic treatment of tumours in the eye, especially retinoblastoma. Phys Med Biol 1982;27:1313–1324.
233. Shields CL, Santos MC, Diniz W, et al. Thermotherapy for retinoblastoma. Arch Ophthalmol 1999;117:885–893.
234. Murray TG, Cicciarelli N, McCabe CM, et al. In vitro efficacy of carboplatin and hyperthermia in a murine retinoblastoma cell line. Invest Ophthalmol Vis Sci 1997;38:2516–2522.
235. Kaneko A. [Malignant ophthalmic tumors]. Nippon Rinsho 1993; 51 Suppl:1013–1020.
236. Murphree AL, Munier FL. Retinoblastoma. In: Ryan SJ, ed. Retina. St. Louis: Mosby, 1994:605–606.
237. Shields CL. Turning up the heat on retinoblastoma. Rev Ophthalmol 1997;4:116–118.
238. Shields JA, Shields CL. Atlas of intraocular tumors. Philadelphia: Lippincott Williams & Wilkins, 1999.
239. Shields JA, Shields CL, de Potter P, et al. Bilateral macular retinoblastoma managed by chemoreduction and chemothermotherapy. Arch Ophthalmol 1996;114:1426–1427.
240. Tucker MA, D’Angio GJ, Boice JD, et al. Bone sarcomas linked to radiotherapy and chemotherapy in children. N Engl J Med 1987;317:588–593.
241. Wong FL, Boice JD, Abramson DH, et al. Cancer incidence after retinoblastoma. JAMA 1997;278:1262–1267.
242. Abramson DH, Ellsworth RM, Kitchin FD, et al. Second nonocular tumors in retinoblastoma survivors. Are they radiation-induced? Ophthalmology 1984;91:1351–1355.
243. Gallie BL, Budning A, DeBoer G, et al. Chemotherapy with focal therapy can cure intraocular retinoblastoma without radiotherapy [published erratum appears in Arch Ophthalmol 1997;115:525] [see comments]. Arch Ophthalmol 1996;114:1321–1328.
244. Shields CL, de Potter P, Himelstein BP, et al. Chemoreduction in the initial management of intraocular retinoblastoma. Arch Ophthalmol 1996;114:1330–1338.
245. Shields CL, Honavar SG, Shields JA, et al. Factors predictive of recurrence of retinal tumors, vitreous seeds, and subretinal seeds following chemoreduction for retinoblastoma. Arch Ophthalmol 2002;120:460–464.
246. Shields CL, Shelil A, Cater J, et al. Development of new retinoblastomas after 6 cycles of chemoreduction for retinoblastoma in 162 eyes of 106 consecutive patients. Arch Ophthalmol 2003;121:1571–1576.
247. Shields CL, Mashayekhi A, Demirci H, et al. Practical approach to management of retinoblastoma. Arch Ophthalmol 2004;122:729–735.
248. Redler LD, Ellsworth RM. Prognostic importance of choroidal invasion in retinoblastoma. Arch Ophthalmol 1973;90:294–296.
249. Pratt CB, Kun LE. Response of orbital and central nervous system metastases of retinoblastoma following treatment with cyclophosphamide/doxorubicin. Pediatr Hematol Oncol 1987;4:125–130.
250. Friedman DL, Himelstein B, Shields CL, et al. Chemoreduction and local ophthalmic therapy for intraocular retinoblastoma. J Clin Oncol 2000;18:12–17.
251. Honavar SG, Singh AD, Shields CL, et al. Does post-enucleation prophylactic chemotherapy in high risk retinoblastoma prevent metastasis? Invest Ophthalmol Vis Sci 2000;41(4):S953.
252. Uusitalo MS, Van Quill KR, Scott IU, et al. Evaluation of chemoprophylaxis in patients with unilateral retinoblastoma with high-risk features on histopathologic examination. Arch Ophthalmol 2001;119:41–48.
253. Chantada G, Fandino A, Davila MT, et al. Results of a prospective study for the treatment of retinoblastoma. Cancer 2004;100:834–842.
254. Abramson DH, Ellsworth RM, Tretter P, et al. Simultaneous bilateral radiation for advanced bilateral retinoblastoma. Arch Ophthalmol 1981;99:1763–1766.
255. Kingston JE, Hungerford JL, Madreperla SA. Results of combined chemotherapy and radiotherapy for advanced intraocular retinoblastoma. Arch Ophthalmol 1996;114:1339–1343.
256. Chan HS, Thorner PS, Haddad G, et al. Multidrug-resistant phenotype in retinoblastoma correlates with P-glycoprotein expression. Ophthalmology 1991;98:1425–1431.
257. Chan HS, Lu Y, Grogan TM, et al. Multidrug resistance protein (MRP) expression in retinoblastoma correlates with the rare failure of chemotherapy despite cyclosporine for reversal of P-glycoprotein. Cancer Res 1997;57:2325–2330.
258. Smith MA, Rubinstein L, Anderson JR, et al. Secondary leukemia or myelodysplastic syndrome after treatment with epipodophyllotoxins. J Clin Oncol 1999;17:569–577.
259. 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;13:902–909.
P.886

260. Kiratli H, Bilgic S, Ozerdem U. Management of massive orbital involvement of intraocular retinoblastoma. Ophthalmology 1998;105:322–326.
261. Pratt CB, Fontanesi J, Chenaille P, et al. Chemotherapy for extraocular retinoblastoma. Pediatr Hematol Oncol 1994;11:301–309.
262. Goble RR, McKenzie J, Kingston JE, et al. Orbital recurrence of retinoblastoma successfully treated by combined therapy. Br J Ophthalmol 1990;74:97–98.
263. Kingston JE, Hungerford JL, Plowman PN. Chemotherapy in metastatic retinoblastoma. Ophthalmic Paediatr Genet 1987;8:69–72.
264. White L. Chemotherapy for retinoblastoma [letter; comment]. Med Pediatr Oncol 1995; 24:341–342.
265. Saleh RA, Gross S, Cassano W, et al. Metastatic retinoblastoma successfully treated with immunomagnetic purged autologous bone marrow transplantation. Cancer 1988;62:2301–2303.
266. Saarinen UM, Sariola H, Hovi L. Recurrent disseminated retinoblastoma treated by high-dose chemotherapy, total body irradiation, and autologous bone marrow rescue. Am J Pediatr Hematol Oncol 1991;13:315–319.
267. Namouni F, Doz F, Tanguy ML, et al. High-dose chemotherapy with carboplatin, etoposide and cyclophosphamide followed by a haematopoietic stem cell rescue in patients with high- risk retinoblastoma: a SFOP and SFGM study. Eur J Cancer 1997;33:2368–2375.
268. Dunkel IJ, Aledo A, Kernan NA, et al. Successful treatment of metastatic retinoblastoma. Cancer 2000;89:2117–2121.
269. Shields JA. Secondary orbital tumors. Diagnosis and management of orbital tumors. Philadelphia: WB Saunders, 1989:341–347.
270. Shields JA, Shields CL, Suvarnamani C, et al. Orbital exenteration with eyelid sparing: indications, technique, and results. Ophthalmic Surg 1991;22:292–297.
271. de Potter P, Shields CL, Shields JA. Clinical variations of trilateral retinoblastoma: a report of 13 cases. J Pediatr Ophthalmol Strabismus 1994;31:26–31.
272. Blach LE, McCormick B, Abramson DH, et al. Trilateral retinoblastoma—incidence and outcome: a decade of experience. Int J Radiat Oncol Biol Phys 1994;29:729–733.
273. Shields CL, Shields JA, Meadows AT. Chemoreduction for retinoblastoma may prevent trilateral retinoblastoma [letter; comment]. J Clin Oncol 2000;18:236–237.
274. Bechrakis NE, Bornfeld N, Schueler A, et al. Clinicopathologic features of retinoblastoma after primary chemoreduction. Arch Ophthalmol 1998;116:887–893.
275. Dithmar S, Rusciano D, Grossniklaus HE. A new technique for implantation of tissue culture melanoma cells in a murine model of metastatic ocular melanoma. Melanoma Res 2000;10:2–8.
276. Mendelsohn ME, Abramson DH, Madden T, et al. Intraocular concentrations of chemotherapeutic agents after systemic or local administration. Arch Ophthalmol 1998;116: 1209–1212.
277. Murray TG, Cicciarelli N, O’Brien JM, et al. Subconjunctival carboplatin therapy and cryotherapy in the treatment of transgenic murine retinoblastoma. Arch Ophthalmol 1997;115:1286–1290.
278. Harbour JW, Murray TG, Hamasaki D, et al. Local carboplatin therapy in transgenic murine retinoblastoma. Invest Ophthalmol Vis Sci 1996;37:1892–1898.
279. Abramson DH, Frank CM, Dunkel IJ. A phase I/II study of subconjunctival carboplatin for intraocular retinoblastoma. Ophthalmology 1999;106:1947–1950.
280. Chevez-Barrios P, Hurwitz MY, Louie K, et al. Metastatic and nonmetastatic models of retinoblastoma. Am J Pathol 2000;157:1405–1412.
281. Mills MD, Windle JJ, Albert DM. Retinoblastoma in transgenic mice: models of hereditary retinoblastoma. Surv Ophthalmol 1999;43:508–518.
282. Trask TW, Trask RP, Aguilar-Cordova E, et al. Phase I study of adenoviral delivery of the HSV-tk gene and ganciclovir administration in patients with current malignant brain tumors. Mol Ther 2000;1:195–203.
283. Hurwitz MY, Marcus KT, Chevez-Barrios P, et al. Suicide gene therapy for treatment of retinoblastoma in a murine model. Hum Gene Ther 1999;10:441–448.
284. Chevez–Barrios P, Chintagumpala MM, Mieler WE, et al. Response of retinoblastoma with vitreous tumor seeding to adenovirus–mediated delivery of thymidine kinase followed by ganciclovir. J Clin Oncol, 2005.