Objective
To study the clinical features and natural history of 17 patients with retinocytoma.
Design
A retrospective case series.
Setting
Tertiary referral center.
Patients
Data on 17 patients with retinocytoma were reviewed for clinical features. The natural history of retinocytoma and its risk for malignant transformation were also evaluated.
Results
Among 920 consecutive patients who had retinoblastoma, retinocytoma, or both, we identified 24 tumors in 17 patients (1.8%) with clinical features compatible with retinocytoma. The median age at diagnosis was 15 years (range, 4-45 years). Of the 24 tumors, the retinocytoma was bilateral in 3 cases (13%) and the family history of retinoblastoma was positive in 3 cases (13%). Seventeen (71%) of the tumors were extramacular in location, and 7 (29%) were located in the macular area. Ophthalmoscopic features characteristic of retinocytoma included the presence of a translucent retinal mass in 21 (88%), calcification in 15 (63%), and retinal pigment epithelial alteration in 13 (54%) of the 24 tumors. A combination of all 3 features was observed in 8 (33%) of the 24 tumors. In 13 (54%) of the tumors, a zone of chorioretinal atrophy could be observed. In 1 patient, subtle tumor regression was documented photographically. Only 1 retinocytoma (4%) underwent malignant transformation into retinoblastoma. At the last follow-up visit, none of the patients had developed a pineoblastoma or another second malignant neoplasm.
Conclusions
Retinocytoma is a rare benign retinal tumor that has characteristic clinical features. The areas of chorioretinal atrophy were suggestive of tumor regression. In our series, the risk for malignant transformation of retinocytoma into retinoblastoma was 4%; therefore, patients with a presumed diagnosis of retinocytoma should be closely observed.
BASED ON clinical, histopathologic, and genetic evidence, it is generally believed that there is a benign variant of retinoblastoma.1-5 The clinical features of these tumors were described in detail by Gallie and associates,1 who suggested that they be called retinoma. Histopathologic studies by Margo and coworkers3 demonstrated that these tumors were composed of well-differentiated, benign-appearing retinal cells. Mitoses and necrosis were characteristically absent, further differentiating these lesions from retinoblastoma. They suggested the alternate term retinocytoma, based on nomenclature used to classify pineal body tumors (benign, pineocytoma; and malignant, pineoblastoma).3 Because retinocytoma resembles irradiated retinoblastoma ophthalmoscopically, other terms such as spontaneously regressed retinoblastoma, spontaneously arrested retinoblastoma, and retinoblastoma group 0 have also been used clinically to describe these tumors.4-6 Overall, retinocytoma may be the preferred terminology because it implies more specifically a benign tumor arising from a retinal cell (retinocyte).
It is important to recognize retinocytoma clinically and differentiate it from active retinoblastoma and other simulating conditions as it usually requires close observation rather than active treatment. However, from a genetic standpoint, retinocytoma is similar to retinoblastoma, with autosomal dominant inheritance involving a mutation in the RB1 gene locus on chromosome 13q14.1,2,5,7,8 In familial cases, various members of a kindred can manifest either retinocytoma or retinoblastoma or even a combination of both between the 2 eyes.1,6,9
In this report, we describe the clinical features and natural history of 17 patients with the clinical diagnosis of retinocytoma. The variations in clinical appearance of retinocytoma and its natural history are interpreted in light of recent advances in the molecular biology of retinoblastoma.
The computerized database of the Oncology Service, Wills Eye Hospital, Philadelphia, Pa, was reviewed, and all patients with retinocytoma, retinoma, or spontaneously arrested or regressed retinoblastoma were retrieved. The patients were diagnosed as having retinocytoma based primarily on ophthalmoscopic appearance and corroborating clinical data, including age at diagnosis, family history, clinical course, and associated ocular findings. Detailed fundus drawings and fundus photographs and descriptions of the retinocytoma were reviewed to document the features of the lesions, such as size, location, presence of a retinal translucent mass, calcification, retinal pigment epithelial alteration, and zone of chorioretinal atrophy. Associated features, such as vitreous calcium deposits, the presence of retinal cysts, retinal traction, and the caliber of retinal feeder vessels, were also noted. Each tumor was classified based on the regression patterns described for irradiated retinoblastoma (type 0, 1, 2, 3, and 4).10
The retinal distribution of the retinocytoma was evaluated by plotting the distance of the epicenter of each lesion from the foveola in clock hour meridians for the left and right eyes separately. The macular area was defined by a circle whose diameter was twice the distance between the optic nerve and the foveola, and the peripheral retina was defined as the part of the retina located anterior to the equator. The intervening zone between the macular area and the equator was designated the midperipheral retina. One clock hour on either side of the vertical axis (the line passing through the 12- to 6-o'clock meridian) and the horizontal axis (the line passing through the 3- to 9-o'clock meridian) was defined as the vertical meridian and the horizontal meridian, respectively.
Information regarding the presence of any second systemic malignant neoplasm and the family history of retinoblastoma were obtained. As necessary, the data were supplemented by directly contacting the patient or referring ophthalmologist. Information obtained at each follow-up visit was evaluated for stability, progression, or regression of the retinocytoma.
Among 920 consecutive patients who had retinoblastoma, retinocytoma, or both on file in the computerized database of the Oncology Service, Wills Eye Hospital, between January 1974 and December 1998, we identified 17 patients (1.8%) who had 24 retinocytomas in 20 eyes (Table 1). The median age at diagnosis of these 17 patients was 15 years (range, 4-45 years). The median follow-up was 49 months (range, 1-206 months). Three patients (18%) had bilateral tumors, and 14 (82%) had unilateral tumors, with 9 (53%) affecting the right eye and 5 (29%) affecting the left eye (Table 2). The diagnosis was made in 7 patients (41%) because of their findings (blurred vision, 4; strabismus, 3; and leukocoria, 0). Seven patients (41%) were asymptomatic. Three patients (18%) were identified when their children developed retinoblastoma (Figure 1). One patient (6%) had retinoblastoma in the other eye. One patient (6%) had retinoblastoma and retinocytoma as 2 separate foci in the same eye.
The salient ophthalmoscopic features of the 24 retinocytomas are summarized in Table 2. The median basal dimension of the tumors was 6.25 mm (range, 0.3-15.0 mm), and the median thickness was 1.75 mm (range, 0.0-5.0 mm). Ophthalmoscopic features included the presence of a translucent retinal mass in 21 (88%), with calcification in 15 (63%) and associated retinal pigment epithelial alteration in 13 (54%), of the 24 tumors (Figure 2). The retinal pigment epithelial alteration usually surrounded the lesion and had well-defined margins. Any 1 of these 3 features was present in all tumors. However, any combination of at least 2 features was present in 17 (71%) of the 24 tumors. All 3 features were present in only 8 (33%) of the 24 tumors. In addition to the triad of features, a surrounding zone of chorioretinal atrophy was also observed in 13 tumors (54%). Including the zone of chorioretinal atrophy as a fourth clinical feature, the diagnostic yield for any combination of at least 2 and 3 features was 83% and 46% (20 and 11 tumors), respectively. Localized calcium deposits were noted in the vitreous in 7 tumors (29%), and retinal traction in 3 (13%). The retinal feeder vessels were of normal caliber in all tumors.
Classification of the ophthalmoscopic appearance of retinocytoma based on retinoblastoma regression patterns revealed the type 0 pattern in 0 tumors (0%), the type 1 pattern in 1 tumor (4%) (Figure 3), the type 2 pattern in 6 tumors (25%) (Figure 4), the type 3 pattern in 14 tumors (58%) (Figure 2), and the type 4 pattern in 3 tumors (13%) (Figure 5).
Tumor foci were distributed in all 3 retinal zones (Table 3). Seven tumors were in the macular area, with none centered in the foveola.
All cases were periodically observed without active treatment. Malignant transformation of retinocytoma into retinoblastoma, characterized by enlargement in size and vitreous seeding, occurred in 1 (4%) of the 24 cases. In this case, stability of retinocytoma was documented for 3 years before transformation into retinoblastoma with tumor enlargement and vitreous seeding, necessitating enucleation.11 In addition, 1 patient subsequently developed a new retinoblastoma in the peripheral retina distant from a posteriorly located retinocytoma. The clinical impression of 2 distinct retinal tumors, a retinocytoma and a retinoblastoma, in the same eye was confirmed histopathologically (Figure 6). In 1 case, subtle progressive regression in the basal dimension of retinocytoma was documented for 206 months (Figure 7). There was no change in the amount of calcification. All remaining tumors remained stable in their ophthalmoscopic appearance. At the last follow-up visit, none of the patients had developed a systemic second malignant neoplasm or symptomatic pineoblastoma.
The reported proportion of retinocytoma among the population with retinoblastoma has varied from 1.8% to 10%.1,9,11 In our series, we observed 17 patients with retinocytoma of a total of 920 patients with retinoblastoma, representing a proportion of 1.8%. The incidence presumably reflects a referral bias as a diagnosis is likely to be made only in symptomatic cases or in those with a family history of retinoblastoma. The diagnosis was made in 10 patients (59%) in our series either on routine eye examination or when the diagnosis of retinoblastoma in another family member prompted an eye examination. Leukocoria, a common initial feature of retinoblastoma, was not an initial feature in any of the 17 patients.
The occurrences of retinocytoma and retinoblastoma are not mutually exclusive; retinocytoma and retinoblastoma have been reported in the same family,1,9,12 in the same patient with retinoblastoma in one eye and retinocytoma in the other,1,9 and as 2 separate foci in the same eye (Figure 6).13 Some cases of retinocytoma are diagnosed when the parents of a child who has retinoblastoma are examined (Figure 1).1,9,12 The diagnosis of an asymptomatic retinocytoma in 1 of the parents of a child with retinoblastoma has major implications in genetic counseling (Figure 1). The examination of first-degree relatives, especially parents, when a new case of retinoblastoma is diagnosed is extremely important.14
The median age at diagnosis was 15 years, and the diagnosis was made in most patients (14 [82%] of 17)after the age of 6 years, in contrast to retinoblastoma, which is usually diagnosed in children younger than 5 years.15
The tumors were randomly distributed in all 3 retinal zones, lacking any preferential location in the retina. In recent studies,16 the retinal distribution of retinoblastoma has been explained based on retinal development. Munier and associates17 reported radial asymmetry with excess foci of retinoblastoma (including 50 retinocytomas) occurring along the horizontal meridian, similar to cone cell topography, suggesting retinoblastoma origin from cone cell lineage. The cell of origin of retinoblastoma is not clearly established because recent histopathologic and immunocytochemical studies18-20 have provided conflicting results suggesting that the retinoblastoma originates from the rod photoreceptors and cone (red-green) photoreceptors.
As reported by others,1,9,12 any 1 of the 3 ophthalmoscopic features of retinocytoma, such as the presence of a translucent retinal mass, calcification, and retinal pigment epithelial alteration, was observed in all retinocytomas (Figure 2). However, any combination of at least 2 features was present in 17 (71%) of the 24 tumors, and all 3 features were present in only 8 (33%) of the tumors. Including a zone of chorioretinal atrophy as a fourth clinical feature,9 the diagnostic yield was higher for any combination of at least 2 and 3 features (83% and 46%, respectively). Localized deposits of calcium in the vitreous, a recently described feature of retinocytoma, were observed in 7 (29%) of the tumors.13 None of the tumors contained intraretinal cysts.21
The ophthalmoscopic appearance of the retinocytomas resembled the spectrum of retinoblastoma regression patterns observed after irradiation. The presence of chorioretinal atrophy in 13 (54%) of the 24 tumors was reminiscent of retinoblastoma regression after irradiation, suggesting tumor regression. The chorioretinal atrophy surrounded the translucent retinal mass and was not limited to the dependent margin of the tumor, unlike the pigmentary changes after a prior retinal detachment. In 1 patient, with the longest follow-up of 206 months, we could document tumor regression photographically (Figure 7). To our knowledge, previous researchers have not documented tumor regression in presumed cases of retinocytoma, and it could be due to lack of photographic documentation, which is necessary to detect minor changes in the tumor dimensions.1 The mechanisms of tumor regression in retinocytoma are unknown but might involve apoptosis.22-24 Regression caused by ischemic or immune-mediated necrosis is incompatible with the histopathologic features of retinocytoma.25
None of the retinocytomas enlarged during observation. Failure to document growth of a retinocytoma may be due to at least 2 reasons. Most retinocytomas may have completed their growth by the time they are diagnosed (after the age of 5 years), and all new lesions found in eyes undergoing prospective examination after conservative treatment of retinoblastoma or due to a family history of retinoblastoma are considered to have retinoblastoma and are treated promptly.
Malignant transformation of retinocytoma into retinoblastoma occurs rarely.7,9,11 We confirmed this histopathologically in 1 (4%) of the 24 tumors. That previously reported case enlarged rapidly and seeded the vitreous at age 7 years after remaining stable for 3 years.11 The results of a histopathologic examination disclosed undifferentiated retinoblastoma overlying a benign retinocytoma consistent with malignant transformation of retinocytoma into retinoblastoma.11 The occurrence of retinoblastoma in adults has been explained based on the malignant transformation of previously undiagnosed asymptomatic retinocytoma.15,26-28 Despite the rarity of malignant transformation, patients with retinocytoma should be observed regularly.
Retinocytoma is considered to be a rare benign phenotypic RB1 gene mutation and carries similar genetic implications as retinoblastoma.1,2,5 Why retinocytoma develops in rare instances rather than the usual phenotype of retinoblastoma is unclear. Hypothetically, retinocytoma could arise if the second hit (M2) occurs at a later stage of cell maturation, when the precursor cell has limited mitotic capability and is unable to sequentially accumulate additional mutations (M3+).7 Retinocytoma could also be a manifestation of low-penetrance retinoblastoma (M1 variation).29-34 In families expressing a low penetrance of retinoblastoma, many gene carriers are either unaffected or affected unilaterally.35,36 Three patients in our series had a family history of retinoblastoma, none of them manifesting evidence of low penetrance. Recent molecular genetic studies29-34,36,37 have attributed low-penetrance retinoblastoma to specific, relatively benign RB1 gene mutations that code for a partially functional retinoblastoma protein. It is possible that the presence of a partially functional retinoblastoma protein in the retinal precursor cell might be able to thwart complete transformation of the retinocyte into retinoblastoma, thereby forming retinocytoma.29-34,36,37
Second malignant neoplasms and the tumors of the pineal region develop in approximately 8% (aged 18 years) and 5% to 8% of patients harboring germline mutations of the RB1 gene, respectively.38,39 None of the patients in our series of retinocytoma developed systemic second malignant neoplasms or pineal tumors. Only 9 patients in our series were putative carriers of a germline mutation (bilateral, familial, and 15% of unilateral cases), and only 1 patient would be expected to develop such tumors. The number of patients in our series is admittedly small; however, review of the large published series1,9,12 of patients with retinocytoma also suggests that second malignant neoplasms are rare in patients with retinocytoma. All 80 patients in a recent review40 of trilateral retinoblastoma had retinoblastoma; none had retinocytoma. It is possible that mechanisms that play a protective role in inducing a benign tumor (retinocytoma) rather than a malignant tumor (retinoblastoma) also protect the extraocular cells from the development of second malignant neoplasms.33,37
In summary, we have provided the clinical features of 17 patients with 24 retinocytomas. In addition to well-recognized features such as the presence of a gray translucent mass, intralesional calcification, and retinal pigment epithelial alteration, the presence of chorioretinal atrophy is an important diagnostic feature. In this large series, nearly 2% of potential retinoblastomas were benign variants. Although most retinocytomas were stable (22 [92%]/24) and demonstrated no tendency to grow or metastasize, patients with retinocytoma should be examined periodically because they can rarely undergo malignant transformation into retinoblastoma.
Accepted for publication August 22, 1999.
This study was supported by the Sarah B. Kant Fund, Philadelphia, Pa (Dr Singh); the Eye Tumor Research Foundation, Philadelphia; Fundação Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Brasilia, Brazil (Dr Santos); the Macula Foundation, New York, NY (Dr C. L. Shields); the Paul Kayser Award of Merit in Retinal Research, Houston, Tex (Dr J. A. Shields); and the Noel T. Simmonds and Sarah L. Simmonds Endowment for Ophthalmic Pathology, Wills Eye Hospital, Philadelphia (Dr Eagle).
Presented in part at the International Ocular Oncology Symposium, Philadelphia, Pa, May 3, 1999; and the Association for Research in Vision and Ophthalmology Meeting, Fort Lauderdale, Fla, May 12, 1999.
Reprints: Arun D. Singh, MD, Oncology Service, Wills Eye Hospital, 900 Walnut St, Philadelphia, PA 19107 (e-mail: arun442@aol.com).
1.Gallie
BAEllsworth
RMAbramson
DHPhillips
RA Retinoma: spontaneous regression of retinoblastoma or benign manifestation of the mutation?
Br J Cancer. 1982;45513- 521
Google ScholarCrossref 2.Gallie
BAPhillips
RAEllsworth
RMAbramson
DH Significance of retinoma and phthisis bulbi for retinoblastoma.
Ophthalmology. 1982;891393- 1399
Google ScholarCrossref 3.Margo
CHidayat
AKopelman
JZimmerman
LE Retinocytoma: a benign variant of retinoblastoma.
Arch Ophthalmol. 1983;1011519- 1531
Google ScholarCrossref 4.Aaby
AAPrice
RLZakov
ZN Spontaneously regressing retinoblastomas, retinoma, or retinoblastoma group 0.
Am J Ophthalmol. 1983;96315- 320
Google Scholar 5.Abramson
DH Retinoma, retinocytoma, and the retinoblastoma gene.
Arch Ophthalmol. 1983;1011517- 1518
Google ScholarCrossref 6.Shields
JAShields
CL Intraocular Tumors: A Text and Atlas. Philadelphia, Pa WB Saunders Co1992;319
7.Gallie
BLDumm
JMChan
HSLHamel
PAPhillips
RA The genetics of retinoblastoma: relevance to the patient.
Pediatr Clin North Am. 1991;38299- 315
Google Scholar 8.Balmer
AMunier
FGailloud
C Retinomes et phtisis bulbi: expression benigne du retinoblastome.
Klin Monatsbl Augenheilkd. 1992;200436- 439
Google ScholarCrossref 10.Singh
ADGarway-Heath
DFPlowman
NKingston
JKHungerford
JL Relationship of regression patterns to recurrence in retinoblastoma.
Br J Ophthalmol. 1993;7712- 16
Google ScholarCrossref 11.Eagle
RC
JrShields
JADonoso
LAMilner
RS Malignant transformation of spontaneously regressed retinoblastoma, retinoma/retinocytoma variant.
Ophthalmology. 1989;961389- 1395
Google ScholarCrossref 12.Lommatzsch
PKZimmermann
WLommatzsch
R Spontane wachstumshemmung beim retinoblastom.
Klin Monatsbl Augenheilkd. 1993;202218- 223
Google ScholarCrossref 13.Lueder
GTHeon
EGallie
BL Retinoma associated with vitreous seeding.
Am J Ophthalmol. 1995;119522- 523
Google Scholar 14.Onadim
ZHykin
PGHungerford
JLCowell
JK Genetic counseling in retinoblastoma: importance of ocular fundus examination of first degree relatives and linkage analysis.
Br J Ophthalmol. 1991;75147- 150
Google ScholarCrossref 17.Munier
FLBalmer
Avan Melle
GGailloud
C Radial asymmetry in the topography of retinoblastoma.
Ophthalmic Genet. 1994;15101- 106
Google ScholarCrossref 18.Nork
TMSchwartz
TLDoshi
HMMillecchia
LL Retinoblastoma: cell of origin.
Arch Ophthalmol. 1995;113791- 802
Google ScholarCrossref 19.Bogenmann
ELochrie
MASimon
MI Cone cell–specific genes expressed in retinoblastoma.
Science. 1987;24076- 78
Google ScholarCrossref 20.Vrabec
TArbizio
VAdamus
G
et al. Rod cell–specific antigens in retinoblastoma.
Arch Ophthalmol. 1989;1071061- 1063
Google ScholarCrossref 21.Morris
WELaPiamna
FG Spontaneous regression of bilateral multifocal retinoblastoma with preservation of normal visual acuity.
Ann Ophthalmol. 1974;61192- 1194
Google Scholar 23.Hetts
SW To die or not to die: an overview of apoptosis and its role in disease.
JAMA. 1998;279300- 307
Google ScholarCrossref 24.Nork
TMPoulsen
GLMillecchia
LLJantz
RGNickells
RW p53 Regulates apoptosis in human retinoblastoma.
Arch Ophthalmol. 1997;115213- 219
Google ScholarCrossref 26.Rychener
RO Retinoblastoma in the adult.
Trans Am Ophthalmol Soc. 1948;46318- 326
Google Scholar 27.Takahashi
TTamura
SInoue
MIsayama
YSashikata
T Retinoblastoma in a 26-year-old adult.
Ophthalmology. 1983;90179- 183
Google ScholarCrossref 28.Nork
TMMillecchia
LLde Venecia
GBMyers
FLVogel
KA Immunocytochemical features of retinoblastoma in an adult.
Arch Ophthalmol. 1996;1141402- 1406
Google ScholarCrossref 29.Onadim
ZHogg
ABaird
PNCowell
JK Oncogenic point mutations in exon 20 of the
RB1 gene in families showing incomplete penetrance and mild expression of the retinoblastoma phenotype.
Proc Natl Acad Sci U S A. 1992;896177- 6181
Google ScholarCrossref 30.Dryja
TPRapaport
JMcGee
TLNork
TMSchwartz
TL Molecular etiology of low-penetrance retinoblastoma in two pedigrees.
Am J Hum Genet. 1993;521122- 1128
Google Scholar 31.Kratzke
RAOtterson
GAHogg
A
et al. Partial inactivation of the RB product in a family with incomplete penetrance of familial retinoblastoma and benign retinal tumors.
Oncogene. 1994;91321- 1326
Google Scholar 32.Schubert
ELStrong
LCHansen
MF A splicing mutation in
RB1 in low penetrance retinoblastoma.
Hum Genet. 1997;100557- 563
Google ScholarCrossref 33.Bremner
RDu
DCConnolly-Wilson
MJ
et al. Deletion of RB exons 24 and 25 causes low penetrance retinoblastoma.
Am J Hum Genet. 1997;61556- 570
Google ScholarCrossref 34.Otterson
GAChen
WCoxon
ABKhleif
SNKaye
FJ Incomplete penetrance of familial retinoblastoma linked to germ-line mutations that result in partial loss of RB function.
Proc Natl Acad Sci U S A. 1997;9412036- 12040
Google ScholarCrossref 35.Macklin
MT A study of retinoblastoma in Ohio.
Am J Ophthalmol. 1960;121- 43
Google Scholar 36.Lohmann
DRBrandt
BHopping
WPassarge
EHorsthemke
B Distinct
RB1 gene mutations with low penetrance in hereditary retinoblastoma.
Hum Genet. 1994;94349- 354
Google ScholarCrossref 37.Sakai
TOhtani
NMcGee
TLRobbins
PDDryja
TP Oncogenic germ-line mutations in Sp1 and ATF sites in the human retinoblastoma gene.
Nature. 1991;35383- 86
Google ScholarCrossref 38.Moll
ACImhof
SMBouter
LMTan
KEWP Second primary tumors in patients with retinoblastoma: a review of the literature.
Ophthalmic Genet. 1997;1827- 34
Google ScholarCrossref 39.De Potter
PShields
CLShields
JA Clinical variations of trilateral retinoblastoma: a report of 13 cases.
J Pediatr Ophthalmol Strabismus. 1994;3126- 31
Google Scholar 40.Marcus
DMBrooks
SELeft
G
et al. Trilateral retinoblastoma: insights into histogenesis and management.
Surv Ophthalmol. 1998;4359- 70
Google ScholarCrossref