Clinical predictors at diagnosis of low risk histopathology in unilateral
advanced retinoblastoma
Stephanie N. Kletke, MD1, Zhao Xun Feng, BSc2, Lili-Naz Hazrati, MD, PhD, FRCPC3, Brenda L. Gallie,
MD, FRCSC1,2,4, Sameh E. Soliman, MD2,5*
Authors’ Affiliations
1 Department of Ophthalmology and Vision Sciences, University of Toronto, Toronto, Canada;
2 Department of Ophthalmology and Vision Sciences, The Hospital for Sick Children, Toronto, Canada;
3 Department of Pediatric Laboratory Medicine, The Hospital for Sick Children, Toronto, Canada;
4 Departments of Molecular Genetics and Medical Biophysics, University of Toronto, Toronto, Canada;
5 Department of Ophthalmology, Faculty of Medicine, Alexandria University, Alexandria, Egypt.
*Corresponding Author: Sameh E. Soliman, 555 University Avenue, Room 7265, Toronto, Canada,
M5G 1X8. [email protected]
Financial Support: None
Conflicts of interest: None
Running Head: Low-risk Histopathology in Unilateral cT2b Retinoblastoma
Number of Figures and Tables: 2 figures, 2 tables and 2 supplementary tables.
Keywords: unilateral retinoblastoma; Group D; histopathology; cancer; primary enucleation; vitreous
seeds.
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
Abstract (348/350)
Objective: Attempted eye salvage for unilateral (cT2b/Group D) retinoblastoma may increase risk of
tumor spread compared to primary enucleation. Identification of clinical features predictive of low
histopathologic risk would support safe trial salvage.
Design: Retrospective, non-comparative single-institutional observational case series
Participants: Children with unilateral cT2b /Group D retinoblastoma (January 2008-February 2018)
who were managed with primary enucleation at the Hospital for Sick Children (SickKids), Toronto,
Canada.
Methods: Data included clinical features (intraocular pressure (IOP), optic nerve obscuration, macular
involvement, tumor seeding and serous retinal detachment (RD) >1 quadrant), timing to enucleation,
histopathological findings, genetic testing and follow-up (metastasis and death).
Main outcome measures: Primary outcome was low-risk (LR) (pT1/pT2) versus high-risk (HR)
(pT3/pT4) (8th Edition American Joint Committee on Cancer) histopathology. Clinicopathologic
correlations were evaluated. Secondary outcome measures were positive predictive values
(probability that certain clinical features would predict LLR histopathology) and, negative
predictive values (probability that absence of these clinical features would predict HLR
histopathology), specificity and sensitivity values.
Results: Thirty-eight eyes were eligible with mean age of 21 months (range: 2-48) at diagnosis. All had
vitreous seeding and normal IOP. The median diagnosis to enucleation interval was 4 days (range, 0–14)
without parental refusals. Histopathology diagnosed 4/38–10.5% HR and 34/38–89.5% LR eyes. HR eyes
demonstrated massive choroidal invasion (4/38–10.5%), or trans-scleral, extraocular and post-laminar
optic nerve invasion (1/38–2.6%). Clinical findings included macular involvement (31/38–82%),
complete optic nerve obscuration (27/38–74%), and RD (28/38–74%). The proportion of eyes with HR
2
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
histopathology was 13% (4/31; 95%CI 1% to 25%) with macular involvement, 15% (4/27; 95%CI 1%
to 28%) with complete optic nerve obscuration, and 14% (4/28; 95%CI 1% to 27%) with RD. The
predictability of LR histopathology was 100% (7/7; 95%CI 47% to 100%) with macular sparing
(7/7), 100% (10/10; 95%CI 63% to 100%) with optic nerve visibility (10/10) and/or 100% (10/10;
95%CI 63% to 100%) with <1 quadrant of RD (10/10) with 100% specificity. One child lacking all
three clinical LR predictive features had HR histopathology (pT3a) developed metastases and died; other
children are alive and well (mean follow-up 65 months).
Conclusion: Presence of macular sparing, optic-nerve visibility and/or <1 quadrant of RD predicted
are highly predictive of LR histopathology at primary enucleation, suggesting that any of these features
predict safe eye salvage. No clinical sign predicted HR histopathology.
3
42
43
44
45
46
47
48
49
50
51
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
Unilateral retinoblastoma eyes staged cT2 (8th Edition American Joint Committee on Cancer (AJCC)
TNMH, tumor, node, metastasis and heritable trait)1 and Group D (International Intraocular
Retinoblastoma Classification, IIRC)2 poses a management challenge. Attempted eye salvage using
primary intra-arterial (IAC)3-7 or systemic chemotherapy8 (both with focal consolidation) is commonly
suggested. However, primary enucleation is an effective option to minimize risk of extraocular extension
and metastasis. The Canadian National Retinoblastoma Strategy Guidelines for Care published in 2009
recommend enucleation of affected unilateral Group D eyes.8
Recently, multiple treatment modalities are suggested to improve success of eye salvage, including
intravitreal chemotherapy (IVitC),9-11 IAC,3-7 periocular chemotherapy,12 and tumor endoresection via pars
plana vitrectomy (PPV).13 The primary concern is whether attempted eye salvage increases the risk of
extraocular tumor dissemination. Our aim was to identify clinical features of primarily enucleated
unilateral cT2 (Group D) eyes that at diagnosis predict low histopathologic risk, in order to guide “safe”
trial eye salvage.
Methods
Study Design
A retrospective, non-comparative, single institutional observational study was conducted in
accordance with the guidelines of the Declaration of Helsinki. Institutional Research Ethics Board
approval was obtained.
Eligibility
Children diagnosed with unilateral Group D retinoblastoma managed with primary enucleation of the
affected eye at the Hospital for Sick Children (SickKids), Toronto, Canada between January 2008
(following submission and implementation of the Canadian guidelines8) through February 2018 were
4
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
evaluated. Exclusion criteria were unilateral retinoblastoma of any other clinical stage, bilaterally affected
children, and cT2 eyes that were secondarily enucleated following trial salvage.
Data Collection
Clinical and Radiological Features
Medical records, including fundus photographs from examinations under anesthesia (EUA), were
reviewed for age at diagnosis and enucleation, laterality, clinical features at presentation (intraocular
pressure (IOP), tumor seeding, optic nerve obscuration (defined as inability to completely see the borders
of the optic nerve head, and subcategorized as “partial” if any portion of the nerve was seen or
“complete” if there was total obscuration), macular involvement and serous retinal detachment (RD),
parental consent to the proposed treatment, eye staging by IIRC, molecular genetic analysis, follow-up
duration, adjuvant treatments received, metastasis and death. Eyes were retrospectively staged by the 8th
Ed. AJCC TNMH.1 Baseline magnetic resonance imaging (MRI) or computed tomography (CT) of the
brain and orbit were reviewed.
Histopathologic Features
Histopathology reports and representative slides were reviewed for all children. Presence of choroidal
invasion was documented as “none”, “focal” (<3 mm) or “massive” (>3 mm in maximum diameter),
based on consensus definitions from the International Retinoblastoma Staging Working Group.14 Invasion
under the retinal pigment epithelium (RPE) but not through Bruch’s membrane was identified. Optic
nerve invasion was categorized as “none”, “prelaminar”, “post-laminar but not to the optic nerve
resection margin” and “tumor at the transected end”.14 Scleral invasion, anterior segment involvement and
extraocular disease were identified. Enucleated eyes were retrospectively staged by the 8th Ed. AJCC
pTNM1 (Table 1).
Outcome Measures
5
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
The primary outcome was the presence of low-risk (LR) histopathology (pT1/pT2) or high-risk (HR)
histopathology (pT3/pT4).1 High-risk histopathologic features included massive choroidal invasion, post-
laminar invasion of the optic nerve head, scleral invasion and extraocular extension.
Clinicopathologic correlation was evaluated. Positive predictive value was the probability of a clinical
feature to predict LR histopathology. Negative predictive value was the probability that absence of this
clinical feature to predict absence of LR (i.e. HR) Histopathology. Sensitivity and specificity for each
clinical sign to predict LR histopathology were calculated.
Statistical Analysis
Results were summarized using frequency/percentage for categorical variables and mean, median,
standard deviation and range for continuous variables. 95% confidence interval for proportions were
calculated using Poisson approximation to binomial distribution for proportions < 0.1 or > 0.9 and normal
approximation method for proportions between 0.1 to 0.9. Groups were compared using Fisher’s exact
test for categorical variables and Student’s t-test for continuous variables. All P-values reported were two-
sided and significance was judged at the 5% level. All analyses were performed using SPSS Version 25
(IBM Corp).
Results
Demographic and Clinical Features
Thirty-eight (Supplementary Table 1) primarily enucleated Group D eyes of 38 children (presenting
age mean 21 months, range 2–48) with unilateral retinoblastoma were included (63% right, 37% left). All
eyes were staged cT2b (8th Edition AJCC).1 There was only one unilateral Group D eye during the study
period not offered enucleation at presentation due to initial staging as Group C. She was offered trial
salvage instead and is excluded from this study. The proper staging was ascertained by the OCT finding
of shallow subclinical detachment that involved the whole inferior retina upstaging it to Group D.15
6
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
At presentation, all eyes had normal IOP. Vitreous seeding was present in all eyes. Tumor involved the
macula in 31/38–82%. Children with macular involvement tended to be younger at diagnosis than
children with macular sparing (mean 20 vs 28 months, respectively) (p=0.09). The optic nerve was
obscured in 28/38–74% (1/28 partial, 27/28 complete obscuration) and RD (>1 quadrant) was present in
28/38–74%. Retinal detachment impaired accurate assessment of subretinal seeding in some eyes.
Concomitant presence of obscured optic nerve and macular involvement was found in 26/38–68%,
obscured optic nerve and RD in 25/38–66%, macular involvement and RD in 28/38–74% and all three
clinical findings in 25/38–66%. Conversely, concomitant finding of visible optic nerve, macular sparing
and <1 quadrant of RD was present in 5/38–13%. The presence of macular involvement, optic nerve
obscuration and RD were positively correlated [macula and optic nerve (p=0.01), macula and RD (p<
0.001), optic nerve and RD (p=0.002)]. Four of 28 eyes with optic nerve obscuration demonstrated
possible optic nerve enhancement on baseline imaging of the brain/orbit. There was no radiological
evidence of extraocular or intracranial involvement in any child.
The median interval from diagnosis to enucleation was 4 days (range, 0–14). Primary enucleation was
performed during the staging EUA for all except one child, for whom enucleation was delayed due to low
partial thromboplastin time. All parents consented to enucleation as the primary treatment.
Histopathologic Features
Choroidal involvement included “none” (26/38–68.4%), “focal” (8/38–21.1%), and “massive” (4/38–
10.5%). Six eyes (15.8%) demonstrated tumour under the RPE without invasion of Bruch’s membrane.
Optic nerve involvement included “none” (10/38–26.3%), “prelaminar invasion” (27/38–71.1%), and
“post-laminar invasion but not to the optic nerve resection margin” (1/38–2.6%). There were no cases of
tumor involvement to the resected margin. One eye (2.6%) demonstrated histopathologic evidence of
anterior segment involvement (pT2b). One eye (2.6%) had trans-scleral and extraocular extension (pT4).
One eye (2.6%) had concomitant massive choroidal invasion, post-laminar optic nerve invasion and
extraocular extension.
7
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
High-Risk Pathology Eyes (Figure 1)
Histopathology review identified 4/38–10.5% HR eyes and 34/38–89.5% LR eyes. HR eyes
demonstrated massive choroidal invasion (4/38–10.5%), and trans-scleral, extraocular and post-laminar
optic nerve invasion (1/38–2.6%). Children with HR versus LR eyes had a higher median age at diagnosis
(31 versus 20 months, p=0.035, Mann-Whitney test). Children <24 months tended to be LR (p=0.04,
Fisher Exact test). The predictability (positive predictive value) of a child <24 months to be LR was
100% (95%CI 82% to 100%), while children ≥24 months to be HR was 22% (95%CI 3% to 41%)
(Table 2). Presenting signs included leukocoria (3/4–75%) and strabismus (1/4–25%). Baseline MRI
brain and orbits showed no evidence of optic nerve, extraocular or intracranial involvement in children
with HR eyes. There was no evidence of metastases at presentation.
Clinicopathologic Correlation
Optic nerve obscuration was not significantly associated with post-laminar optic nerve invasion in this
cohort (p=1.000). Macular involvement was not significantly associated with massive choroidal invasion
(p=0.557) or scleral invasion (p=1.000). Serous RD was not significantly associated with massive
choroidal invasion (p=0.287) or scleral invasion (p=1.000, Supplementary Table 2). None of the eyes
showing enhanced optic nerve on MRI at presentation had post-laminar nerve invasion (p=1.000). All of
the 13 eyes with at least one of one of the following three signs: of macular sparing, optic nerve
visibility or <1 quadrant of RD showed LR histopathology. Twenty-one eyes (84%, 21/25) lacked all 3
signs and showed LR histopathology (p=0.28) (Figure 2). The predictability (positive predictive value)
that an eye had LR histopathology was 100% (95%CI 47% to 100%) with macular sparing, 100%
(95%CI 63% to 100%) with optic nerve visibility and, 100% (95%CI 63% to 100%) with <1 quadrant
of RD. and Eyes withany combination of two or three of these clinical findings all was also 100%
predictive ofhad LR histopathology with 100% specificity (i.e. truly have LR histopathology) (Table 2).
The negative predictive value that an eye lack LR histopathology (i.e. HR histopathology) was 13%
(95%CI 1% to 25%) with macular involvement, 14% with any degree of optic nerve obscuration
8
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
(95%CI 1% to 27%), 15% (95%CI 1% to 28%) with complete optic nerve obscuration and 14%
(95%CI 1% to 27%) with RD. The predictability of HR histopathology with concomitant findings of
obscured optic nerve and macular involvement was 15% (95%CI 2% to 29%), obscured optic nerve
and RD was 16% (95%CI 2% to 30%), macular involvement and RD was 14% (95%CI 1% to 27%),
and all three clinical findings was 16% (95%CI 2% to 30%). The sensitivity (i.e. rule out LR
histopathology) was low (table 2).
Molecular Analysis
Molecular genetic testing was performed on tumor samples from 37 enucleated eyes while one eye
(HX) was untested. The two tumor RB1 pathogenic variants were identified in 34/37 eyes. Blood tests for
these pathogenic variants showed that the patients were H116 (7/38–18%) or H0* (27/38–71%); 3 patients
were HX (heritability cannot be verified). The H1 children all had known RB1 pathogenic variants
associated with unilateral disease: mosaicism (3/7), low penetrance RB1 pathogenic variant (3/7) and 13q
deletion syndrome (1/7). The children that showed high-risk pathology were H1 (1 pT4 eye, Figure 1A),
H0*16 (pT3a, Figure 1B) and HX (2 pT3a eyes, one died, Figure 1C, D).
Follow-up, Metastasis and Death
At mean follow-up of 65 months, one child (2.6%) lacking all three clinical LR predictive features
showed HR histopathology (pT3a, massive choroidal invasion) and developed metastases and died. Bony
metastases were found 1 year following retinoblastoma diagnosis.17 The child received six cycles of
systemic chemotherapy, autologous bone marrow transplant and focal radiation. While ocular pathology
was initially interpreted as LR, internal retrospective review identified a previously not seen area of
massive choroidal invasion. Metastatic surveillance remained negative until 1 year later, when
intracranial dural metastases were identified on MRI. The child died 18 months after metastasis diagnosis.
The other children in this cohort are alive and well. No child was lost to follow-up.
9
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
Discussion
The International Intraocular Retinoblastoma Classification (IIRC) (2005) staged eyes clinically
Group A (very low risk) through E (very high risk) to predict outcomes following systemic
chemoreduction and focal therapy.2 The 8th Edition TNMH was based on evidence from an international
survey,1 now recommended for retinoblastoma staging. Advanced intraocular disease includes Group E
eyes [with phthisis bulbi (cT3a), anterior segment tumor invasion (cT3b), rubeosis irides with neovascular
glaucoma (cT3c), hyphema and/or massive vitreous hemorrhage (cT3d) and aseptic orbital cellulitis
(cT3e)], and Group D eyes [with significant RD (cT2a) and/or seeding (vitreous and/or subretinal, cT2b)].
High-risk histopathologic features of the enucleated eye predictive of increased metastatic risk are
pT3/pT4, which include massive choroidal invasion,18,19 post-laminar invasion of the optic nerve head
with or without a positive margin,18,20 scleral invasion and extraocular extension (Table 1).1
Isolated massive choroidal invasion is controversial as a risk factor for metastasis. Shields et al.19
found that patients with choroidal invasion were more likely to develop metastases, but this was only
significant in the presence of concomitant optic nerve invasion. While some groups report significantly
reduced metastatic risk with postenucleation adjuvant therapy for massive choroidal invasion,21 others
report good prognosis without adjuvants.22,23 For isolated massive choroidal invasion, Kim24 estimated an
extraocular relapse rate of 5.5% versus 0% without and with adjuvant therapy, respectively. Similar
survival rates for focal versus massive choroidal invasion have been reported without adjuvants, provided
effective treatments for relapse are available.25 Toxicity associated with systemic adjuvant therapy
requires consideration during decision-making.26 In our center, massive choroidal invasion is considered a
HR feature requiring adjuvant chemotherapy.8
The first goal of treatment for advanced unilateral retinoblastoma is to save the child’s life and prevent
extraocular tumour dissemination; secondary goal is to save a seeing eye. Primary enucleation is a safe
and cheap option, allowing an early return to normal life,27 fewer interventions and EUAs,28 less
10
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
socioeconomic impact,29 and histopathologic review of disease extent to guide further therapy. Primary
enucleation is current practice for cT3 (Group E) eyes and many cT2b (Group D) eyes (Group D). In our
cohort all parents accepted our recommendation for primary enucleation. Parental acceptance of
enucleation is influenced by physician education and counselling.30 Domingo et al. studied the
psychosocial factors affecting parental decision-making regarding enucleation and suggested that
“explanation on the course and prognosis of retinoblastoma may make the attitude of the parents more
favorable toward seeking surgery”.31
Potential for useful vision is important but salvage of a unilateral blind eye for cosmesis may not be
justified, given good prosthesis movement with myoconjunctival enucleation.32 Prolonged attempts at
globe salvage may delay diagnosis of HR features exposing the child to risk of metastasis. Pre-
enucleation chemotherapy may downstage pathological findings, delay enucleation and obstruct
recognition of HR disease.33 With no randomized controlled trial evidence to guide management of cT2
eyes, the clinician and family can balance the impact of potential years of trial salvage with hidden risks,33
against vision potential and quality of life for the child and family. Systemic chemotherapy, IAC, IVitC,
periocular chemotherapy and PPV are available for attempted eye salvage. The success of IVitC to
control vitreous disease9-11 justifies trial salvage for a unilateral cT2b eye, as long as metastatic risk is
minimal. The incidence of metastasis post-IAC is vague,5 as 3 metastatic deaths34 were reported as a sole
outcome for the collective data of 6 centers on 1177 eyes of 1139 patients without reporting how many
children had metastasis.35,36 This management debate clearly shows the value of predictors of either HR or
LR histopathology at diagnosis to facilitate decision making.
Clinical features at presentation reported to predict HR histopathology predominantly describe cT3
(Group E) eyes, not cT2b eyes, and include older age, symptoms >6 months, hyphema, pseudohypopyon,
orbital cellulitis, secondary glaucoma and buphthalmos.37-39 Furthermore, exophytic growth pattern, tumor
thickness >15 mm and vitreous hemorrhage predict optic nerve invasion,20 and iris neovascularization is
associated with choroidal invasion.19,40 Yousef et al.41 concluded that clinical staging alone (TNM 7th ed.,
11
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
IIRC or Reese Ellsworth classification) is insufficient to predict HR histopathology. This encouraged us
to study individual clinical findings as predictors of HR/LR histopathology. Useful predictive,
reproducible clinical findings would be those easily identified at initial staging EUA, the time of decision
of trial salvage versus primary enucleation. We excluded subjective findings such as presenting
complaint, duration of symptoms, pattern of growth (endophytic, exophytic or mixed) and presence of
subretinal seeding under detached retina. The value of MRI in predicting HR histopathology is
minimal.42,43
Approximately 2–33% of Group D eyes are expected to harbour HR histopathologic features in
primary enucleated eyes.27,39,41,44-47 However, the heterogeneous literature is limited by non-consensus in
defining HR histopathology features, variable classifications and inclusion of primarily and secondarily
enucleated eyes and children with unilateral and bilateral disease, and eyes with no staging recorded. In
our cohort of primarily enucleated unilateral cT2b eyes, 10.5% had HR histopathology. Macular
involvement, optic nerve obscuration and >1 quadrant of RD each was not helpful to predict HR
histopathology (predictive value 13%, 14%, or 14%, respectively). However, presence of macular
sparing, visible optic nerve and <1 quadrant of RD werewas each and any of their combinations
highly predictive of LR histopathology. with high specificity (100%). This suggests that such eyes are
appropriate for cautious trial salvage, especially if all three findings are present.
Fabian et al.46 reported on 40 primarily enucleated IIRC Group D eyes (all cT2b, 37 unilateral). At
presentation, 95% (38/40) had foveal involvement, 95% (38/40) had optic disc obscuration and 97% had
RD, compared to 82%, 74% and 74%, respectively in our cohort. They reported absence of vitreous seeds
as the sole significant predictor of HR based on p=0.042,46 whereas all eyes in our study had vitreous
seeds. Small sample sizes render tests of significance inaccurate as one extra entry can shift the p-value
significantly.
We used predictive values as well as significance tests to interpret our data. Significance tests showed
non-significant associations with wide confidence intervals due to the small sample size and low rate of
12
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
positive events. When we combined our data with Fabian et al.46 (78 eyes, Table 2), absence of vitreous
seeds was not significant (p=0.05) and other factors showed a lower p-value than reported by others.
When we applied predictive values, we showed visible optic nerve was 100% (95%CI 69% to 100%)
predictive of LR histopathology and non-involved fovea was 100% (95%CI 59% to 100%) predictive of
LR histopathology. we showed 100% predictive of LR histopathology with visible optic nerve and non-
involved fovea. Absence of vitreous seeds showed 71% (95%CI 48% to 95%) predictability of having
LR histopathology. Berry et al.47 reported that at diagnosis 15% of eyes with optic nerve obscuration
(69/102), and 0% with visible optic nerve (33/102) had post-laminar invasion following primary
enucleation, suggesting a possible clinicopathologic association. This is in accordance with 100%the
high predictability for LR histopathology with visible optic nerve in our cohort. The degree of optic
nerve obscuration could not be assessed as a predictor given only one eye demonstrated partial nerve
obscuration.
The mean age for unilateral retinoblastoma is 24 months.16 Children diagnosed at a younger age
(<24 months) showed 100% (95%CI 82% to 100%) predictability of LR eyes in our cohort. This is
probably due to earlier diagnosis at a less advanced intraocular stage. Fabian et al.46 showed similar
results after excluding the one HR group D eye diagnosed at 1 month of age as it was in the context of
bilateral retinoblastoma. Age should be treated cautiously as a clinical predictor since the subjective
duration of symptoms and delay in diagnosis might confound its value.
Our proposed clinical predictors are at initial diagnosis and there is no evidence to suggest their
usefulness in decision-making for recurrent or refractory cases. These predictors do not predict outcomes
as LR eyes at diagnosis are not guaranteed to remain LR if unresponsive to initial treatment. The decision
of extended duration trial salvage may depend on realistic outcomes including metastatic risks, treatment
morbidity, socioeconomic impact and visual potential taking into consideration that the other eye is
perfectly normal.
13
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
The limitations of this study include its retrospective design and relatively small sample size.
However, our stringent inclusion criteria of only unilateral, primarily enucleated Group D eyes achieved a
homogenous study population. Another limitation was the low rate of positive events, which limits
statistical analysis of associations between clinical findings and histopathologic features of the included
eyes. Furthermore, our analysis is relevant only to the point of diagnosis and does not address progress of
histopathologic risk with extended treatments.
Conclusion
Macular sparing, optic nerve visibility and <1 quadrant of RD at presentation were highly predictive of
LR, and may predict which advanced eyes are suitable for trial salvage decision. HR histopathology was
found in 10.5% of primarily enucleated unilateral cT2b (Group D) eyes. No clinical findings were found
to be predictive of HR histopathology. Previous reports have demonstrated acceptable salvage rates in
unilateral Group D eyes. We recommend taking into consideration the previous aforementioned factors
before undertaking a decision for more cautious trial salvage in eyes with macular involvement, optic
nerve obscuration and >1 quadrant RD. Debates on management of unilateral cT2 eyes would be solved
by robust, multicenter collaborative studies involving many children to establish clinicopathologic
correlations and multifaceted outcomes.
References
1. Mallipatna A, Gallie BL, Chévez-Barrios P, et al. Retinoblastoma. In: Amin MB, Edge SB,
Greene FL, eds. AJCC Cancer Staging Manual. Vol 8th Edition. New York, NY: Springer;
2017:819-831.
2. Murphree AL. Intraocular retinoblastoma: the case for a new group classification.
Ophthalmology clinics of North America. 2005;18:41-53.
14
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
3. Munier FL, Mosimann P, Puccinelli F, et al. First-line intra-arterial versus intravenous
chemotherapy in unilateral sporadic group D retinoblastoma: evidence of better visual outcomes,
ocular survival and shorter time to success with intra-arterial delivery from retrospective review
of 20 years of treatment. Br J Ophthalmol. 2016.
4. Shields CL, Manjandavida FP, Lally SE, et al. Intra-arterial chemotherapy for retinoblastoma in
70 eyes: outcomes based on the international classification of retinoblastoma. Ophthalmology.
2014;121(7):1453-1460.
5. Yousef YA, Soliman SE, Astudillo PP, et al. Intra-arterial Chemotherapy for Retinoblastoma: A
Systematic Review. JAMA ophthalmology. 2016;134(6):584-591.
6. Gobin YP, Dunkel IJ, Marr BP, Brodie SE, Abramson DH. Intra-arterial chemotherapy for the
management of retinoblastoma: four-year experience. Arch Ophthalmol. 2011;129(6):732-737.
7. Suzuki S, Yamane T, Mohri M, Kaneko A. Selective ophthalmic arterial injection therapy for
intraocular retinoblastoma: the long-term prognosis. Ophthalmology. 2011;118(10):2081-2087.
8. Canadian Retinoblastoma S. National Retinoblastoma Strategy Canadian Guidelines for Care:
Strategie therapeutique du retinoblastome guide clinique canadien. Can J Ophthalmol. 2009;44
Suppl 2:S1-88.
9. Munier FL, Gaillard MC, Balmer A, et al. Intravitreal chemotherapy for vitreous disease in
retinoblastoma revisited: from prohibition to conditional indications. Br J Ophthalmol.
2012;96(8):1078-1083.
10. Munier FL, Soliman S, Moulin AP, Gaillard MC, Balmer A, Beck-Popovic M. Profiling safety of
intravitreal injections for retinoblastoma using an anti-reflux procedure and sterilisation of the
needle track. Br J Ophthalmol. 2012;96(8):1084-1087.
11. Berry JL, Shah S, Bechtold M, Zolfaghari E, Jubran R, Kim JW. Long-term outcomes of Group
D retinoblastoma eyes during the intravitreal melphalan era. Pediatr Blood Cancer. 2017.
15
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
12. Mallipatna AC, Dimaras H, Chan HS, Heon E, Gallie BL. Periocular topotecan for intraocular
retinoblastoma. Arch Ophthalmol. 2011;129(6):738-745.
13. Zhao J, Li Q, Wu S, et al. Pars Plana Vitrectomy and Endoresection of Refractory Intraocular
Retinoblastoma. Ophthalmology. 2018;125(2):320-322.
14. Sastre X, Chantada GL, Doz F, et al. Proceedings of the consensus meetings from the
International Retinoblastoma Staging Working Group on the pathology guidelines for the
examination of enucleated eyes and evaluation of prognostic risk factors in retinoblastoma.
Archives of pathology & laboratory medicine. 2009;133(8):1199-1202.
15. Soliman SE, VandenHoven C, MacKeen LD, Heon E, Gallie BL. Optical Coherence
Tomography-Guided Decisions in Retinoblastoma Management. Ophthalmology.
2017;124(6):859-872.
16. Soliman SE, Racher H, Zhang C, MacDonald H, Gallie BL. Genetics and Molecular Diagnostics
in Retinoblastoma--An Update. Asia Pac J Ophthalmol (Phila). 2017;6(2):197-207.
17. Racher H, Soliman S, Argiropoulos B, et al. Molecular analysis distinguishes metastatic disease
from second cancers in patients with retinoblastoma. Cancer Genet. 2016;209(7-8):359-363.
18. 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(6):1206-1213.
19. Shields CL, Shields JA, Baez KA, Cater J, De Potter PV. Choroidal invasion of retinoblastoma:
metastatic potential and clinical risk factors [see comments]. British Journal of Ophthalmology.
1993;77(9):544-548.
20. Shields CL, Shields JA, Baez K, Cater JR, De-Potter P. Optic nerve invasion of retinoblastoma.
Metastatic potential and clinical risk factors. Cancer. 1994;73(3):692-698.
16
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
21. Honavar SG, Singh AD, Shields CL, et al. Postenucleation adjuvant therapy in high-risk
retinoblastoma. Arch Ophthalmol. 2002;120(7):923-931.
22. Uusitalo MS, Van Quill KR, Scott IU, Matthay KK, Murray TG, O'Brien JM. Evaluation of
chemoprophylaxis in patients with unilateral retinoblastoma with high-risk features on
histopathologic examination. Arch Ophthalmol. 2001;119(1):41-48.
23. Chantada GL, Dunkel IJ, de Davila MT, Abramson DH. Retinoblastoma patients with high risk
ocular pathological features: who needs adjuvant therapy? Br J Ophthalmol. 2004;88(8):1069-
1073.
24. Kim JW. Retinoblastoma: evidence for postenucleation adjuvant chemotherapy. Int Ophthalmol
Clin. 2015;55(1):77-96.
25. Bosaleh A, Sampor C, Solernou V, et al. Outcome of children with retinoblastoma and isolated
choroidal invasion. Arch Ophthalmol. 2012;130(6):724-729.
26. Perez V, Sampor C, Rey G, et al. Treatment of Nonmetastatic Unilateral Retinoblastoma in
Children. JAMA ophthalmology. 2018;136(7):747-752.
27. Mallipatna AC, Sutherland JE, Gallie BL, Chan H, Heon E. Management and outcome of
unilateral retinoblastoma. J AAPOS. 2009;13(6):546-550.
28. Fabian ID, Stacey AW, Johnson KC, et al. Primary enucleation for group D retinoblastoma in the
era of systemic and targeted chemotherapy: the price of retaining an eye. Br J Ophthalmol. 2017.
29. Soliman SE, Dimaras H, Souka AA, Ashry MH, Gallie BL. Socioeconomic and psychological
impact of treatment for unilateral intraocular retinoblastoma. Journal Francais D Ophtalmologie.
2015;38:550—558.
30. Olteanu C, Dimaras H. Enucleation Refusal for Retinoblastoma: A Global Study. Ophthalmic
Genet. 2014:1-7.
17
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
31. Domingo RED, Toledo MSW, Mante BVL. Psychosocial Factors Influencing Parental Decision
to Allow or Refuse Potentially Lifesaving Enucleation in Children with Retinoblastoma. Asia
Pac J Oncol Nurs. 2017;4(3):191-196.
32. Shome D, Honavar SG, Raizada K, Raizada D. Implant and prosthesis movement after
enucleation: a randomized controlled trial. Ophthalmology. 2010;117(8):1638-1644.
33. Zhao J, Dimaras H, Massey C, et al. Pre-enucleation chemotherapy for eyes severely affected by
retinoblastoma masks risk of tumor extension and increases death from metastasis. Journal of
clinical oncology : official journal of the American Society of Clinical Oncology. 2011;29(7):845-
851.
34. Abramson DH, Shields CL, Jabbour P, et al. Metastatic deaths in retinoblastoma patients treated
with intraarterial chemotherapy (ophthalmic artery chemosurgery) worldwide. Int J Retina
Vitreous. 2017;3:40.
35. Soliman SE, Dimaras H, Gallie B, Shaikh F. Re: Metastatic deaths in retinoblastoma patients
treated with intraarterial chemotherapy (ophthalmic artery chemosurgery) worldwide. Int J
Retina Vitreous. 2018;4:19.
36. Soliman SE, Gallie BL, Shaikh F. Intra-arterial Chemotherapy for Retinoblastoma-Reply. JAMA
ophthalmology. 2016;134(10):1203.
37. Kashyap S, Meel R, Pushker N, et al. Clinical predictors of high risk histopathology in
retinoblastoma. Pediatr Blood Cancer. 2012;58(3):356-361.
38. Chantada GL, Gonzalez A, Fandino A, et al. Some clinical findings at presentation can predict
high-risk pathology features in unilateral retinoblastoma. J Pediatr Hematol Oncol.
2009;31(5):325-329.
39. Kaliki S, Srinivasan V, Gupta A, Mishra DK, Naik MN. Clinical Features Predictive of High-
Risk Retinoblastoma in 403 Asian Indian Patients: A Case-Control Study. Ophthalmology. 2015.
18
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
40. Chawla B, Sharma S, Sen S, et al. Correlation between clinical features, magnetic resonance
imaging, and histopathologic findings in retinoblastoma: a prospective study. Ophthalmology.
2012;119(4):850-856.
41. Yousef YA, Al-Hussaini M, Mehyar M, et al. Predictive Value of Tnm Classification,
International Classification, and Reese-Ellsworth Staging of Retinoblastoma for the Likelihood of
High-Risk Pathologic Features. Retina. 2015.
42. Hiasat JG, Saleh A, Al-Hussaini M, et al. The predictive value of magnetic resonance imaging of
retinoblastoma for the likelihood of high-risk pathologic features. Eur J Ophthalmol.
2018:1120672118781200.
43. de Jong MC, de Graaf P, Noij DP, et al. Diagnostic performance of magnetic resonance imaging
and computed tomography for advanced retinoblastoma: a systematic review and meta-analysis.
Ophthalmology. 2014;121(5):1109-1118.
44. Wilson MW, Qaddoumi I, Billups C, Haik BG, Rodriguez-Galindo C. A clinicopathological
correlation of 67 eyes primarily enucleated for advanced intraocular retinoblastoma. The British
journal of ophthalmology. 2011;95(4):553-558.
45. Kaliki S, Tahiliani P, Mishra DK, Srinivasan V, Ali MH, Reddy VA. OPTIC NERVE
INFILTRATION BY RETINOBLASTOMA: Predictive Clinical Features and Outcome. Retina.
2016;36(6):1177-1183.
46. Fabian ID, Stacey AW, Chowdhury T, et al. High-Risk Histopathology Features in Primary and
Secondary Enucleated International Intraocular Retinoblastoma Classification Group D Eyes.
Ophthalmology. 2017;124(6):851-858.
47. Berry JL, Kogachi K, Jubran R, Kim JW. Loss of fundus view as an indication for secondary
enucleation in retinoblastoma. Pediatr Blood Cancer. 2017.
19
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
Tables
Table 1. American Joint Committee on Cancer (AJCC) pathological staging 8th Edition.
Table 2. Predictive values versus significance (p-value) in analysis of Fabian et al 46 and current sample.
Supplementary Table 1. Clinicopathologic features of the studied sample (n=38).
Supplementary Table 2. Significance of association between clinical findings and histopathologic features
of enucleated eyes.
Figure Legend
Figure 1. (A) Left, wide-field fundus image showing a right multilobulated tumor with overlying
serous retinal detachment (RD) and subretinal seeding with obscured optic nerve and macular
involvement with tumor. Middle left, Histopathological section under low magnification through the optic
nerve demonstrating extra-scleral (yellow-box) and post-lamina cribrosa invasion (green-box), but not to
the optic nerve resection margin (pT4). Middle right, high magnification of trans-scleral and extra-scleral
invasion (arrowheads). Right, High magnification showing post-laminar invasion (arrowheads). Whole-
body MRI (WBMRI), lumbar puncture (LP) and bilateral bone marrow aspirate (BMA) were negative for
malignancy. He underwent six cycles of vincristine, etoposide, carboplatin (VEC) and cyclophosphamide,
followed by orbital irradiation. Child is alive and disease free after 3 years follow-up.
(B) Left, wide-field fundus image demonstrating a large inferior tumor with overlying RD, vitreous
and subretinal seeds. The optic nerve was obscured and the macular involved with tumor. Middle and
right, Histopathological sections under low and intermediate magnification showing massive choroidal
invasion (arrowheads) beyond the confines of the retinal pigment epithelium (RPE), with no evidence of
scleral invasion (pT3a). There was prelaminar optic nerve invasion. LP and BMA were negative for
malignancy and she received four cycles of VEC. The child is alive and disease free after 3 years follow
up.
20
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
Soliman et al. Low-risk Histopathology in Unilateral cT2b Retinoblastoma
(C) Left, wide-field fundus image of demonstrating a large tumor obscuring the nerve and macula,
with associated hemorrhage, RD and diffuse vitreous seeding. There was no anterior segment extension
evident on ultrasound biomicroscopy. Histopathology was confirmed to be massive choroidal invasion
(arrowheads) (pT3a). The child received systemic adjuvant chemotherapy and alive after 29 months
follow up.
(D) Left, wide-field fundus image demonstrating a large tumor with associated RD, subretinal and
focal vitreous seeding. There was no visualization of the optic nerve and the macula was involved. Initial
review was consistent with low-risk histopathology. One year later the child presented with fever and
pain, and WBMRI identified a paraspinal tumor. Molecular analysis confirmed metastasis.17 Bilateral
BMA were involved with tumor cells. MRI showed no orbital or intracranial disease and LP was
negative. Internal review of the ocular pathology, including further choroidal sections, showed an area of
massive choroidal invasion (yellow-box) (pT3a). The child received 6 cycles of VEC and cyclosporine,
followed by autologous bone marrow transplant and focal irradiation. He was diagnosed with dural-based
metastases 1 year later. Despite radiotherapy, the child died 18 months after presentation with metastases.
LC– lamina cribrosa, S – sclera, RPE – retinal pigment epithelium, arrowheads denotes tumor
invasion boundaries, blue boxes denotes highlighted areas massive choroidal and/or scleral invasion,
green box denotes highlighted areas of optic nerve invasion
Figure 2: Schematic presentation of the predictive clinical signs of low-risk histopathology and their
clinicopathologic correlation.
21
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471