Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 1
Predictive factors for Retrospective analysis of clinical and genetic associations
for Carboplatin related ototoxicity in children treated for Retinoblastoma
Crystal N D’Silva1* | Sameh E. Soliman2,3* | Helen Dimaras2,4,5,6 | Irakli Dzenladze1| Helen
Chan7 | Brenda L. Gallie1,2,4, §
1Department of Medical Biophysics, University of Toronto, Toronto, Canada
2Department of Ophthalmology and Vision Science, Hospital for Sick Children, Toronto,
Canada
3Department of Ophthalmology, Faculty of Medicine, University of Alexandria,
Alexandria, Egypt
4Department of Ophthalmology and Vision Science, University of Toronto, Toronto,
Canada
5Child Health Evaluative Sciences Program, SickKids Research Institute, Toronto,
Canada
6Division of Clinical Public Health, Dalla Lana School of Public Health, University of
Toronto, Toronto, Canada
7Division of Haematology/Oncology, Hospital for Sick Children, Toronto, Canada
*These authors contributed equally to this work.
§Corresponding author: Dr. Brenda Gallie at the Department of Ophthalmology and
Vision Sciences, the Hospital for Sick Children, 555 University Avenue, Toronto, ON
M5G 1X8, Canada. Email: [email protected]. Tel.: (+1) 416-294-9729, Fax 1-866-833-
5157)
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 2
A Research Article submitted to Pediatric Blood and Cancer.
This work was partially presented as a paper at the Association for Research in Vision
and Ophthalmology, Seattle, Washington, 5 May 2016.
Word Count: Abstract (247/250), Main text (2336/3500)
Tables and Figures: 4 tables and 2 figures
Supplemental material: 1 file, 4 tables and 2 figures
Running Title: Carboplatin ototoxicity in retinoblastoma patients
KEYWORDS ototoxicity, carboplatin, retinoblastoma, cancer, genetics, chemotherapy
Abbreviations: IIRC, International Intraocular Retinoblastoma classification; AUC, area
under the curve; TPMT, Thiopurine S-methyltransferase; SIOP, International Society of
Paediatric Oncology; COMT, Catechol-O-methyltransferase; CCG, Children Cancer
Group; ABCC3, ATP-binding cassette sub-family C member 3; NCI-CTCAE, National
Cancer Institute Common Terminology Criteria for Adverse Events; ROC, Receiver
operating characteristic; CEV, carboplatin, etoposide, vincristine; CTV, carboplatin,
teniposide, vincristine
2
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 3
Abstract: (247240/250)
Background: Children with retinoblastoma treated with carboplatin-based
chemotherapy are at risk of moderate to severe, irreversible hearing loss. Based on
published evidence, we hypothesized that risk for ototoxicity is associated with clinical
parameters and variants in candidate genes in drug metabolism pathways
(methyltransferases TPMT and COMT, and drug transporter ABCC3), would be
predictive of ototoxicity.
Procedure: Retrospective review of clinical records of retinoblastoma patients treated
with carboplatin-based chemotherapy recorded age (at diagnosis and chemotherapy
initiation), chemotherapy sessions (cycles number, drug doses and cumulative
carboplatin dose), and hearing loss (defined as ototoxicity ≥ grade 2 by at least one
classification system). Blood samples were genotyped for genetic variants in TPMT
(rs12201199, rs1800460), COMT (rs4646316, rs9332377), and ABCC3 (rs1051640) by
quantitative PCR and confirmed by allele-specific PCR.
Results: Full audiometric data and Stored stored DNA of 97 retinoblastoma patients
was available with full audiometric data were available for 71 retinoblastoma patients
who were included (88% carried a RB1 pathogenic variant allele). Median carboplatin
cumulative dose was 1400 mg/m2 (260-5148 mg/m2). Ototoxicity occurred in 18 patients
(25%), significantly associated with age at diagnosis (p=0.01) and age at chemotherapy
initiation (OR=4.99, p=0.008). The highest likelihood ratio of hearing loss was
associated with chemotherapy initiation <4.25 months of age. Ototoxicity was not
associated with any tested genetic variants.
3
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 4
Conclusions: We observed a 25% incidence of ototoxicity in retinoblastoma patients
treated with carboplatin, higher than previously published figures. Age at treatment
initiation was a risk predictor ofassociated with carboplatin-induced ototoxicity, with
children <4.25 months of age at highest risk. None of the studied genetic variants was
associated with ototoxicity.
4
62
63
64
65
66
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 5
1 | INTRODUCTION
Retinoblastoma, the most common ocular childhood cancer, originates from a
developing retinal cell in one or both eyes.1 All bilateral (100%) and 15% of unilateral
retinoblastoma carry RB1 germline mutations.2 Treatment is determined by age at
diagnosis, laterality, stage of disease at presentation and overall cancer staging.
Treatment regimens encompass different types of chemotherapy, laser therapy,
cryotherapy, radiotherapy and surveillance.3,4
Platinum-induced ototoxicity presents as bilateral high-frequency sensorineural
hearing loss, with increasing incidence and severity in response to cumulative dosage.5,6
Hearing impairment has negative effects on quality of life, cognitive development, and
learning, particularly in paediatric patients7-9 who may have a concurrent degree of
visual impairment from retinoblastoma or its treatment.
Carboplatin is a platinum-based chemotherapeutic agent used in most systemic
chemotherapy protocols for retinoblastoma treatment10-14. While carboplatin is less toxic
than cisplatin15,16, the reported rate of carboplatin related ototoxicity occurrence highly
varies from 0% (no hearing impairment) 17-22 to 16.7 %23-25. Cumulative carboplatin
doses, younger age at treatment initiation, and radiation therapy were identified as
potential risk factors for sustained hearing loss post-treatment.24-26
there have been reports of ototoxicity ranging from 0.009% to 16.7% of paediatric
patients treated with carboplatin for retinoblastoma.12-14 In contrast, other studies report
no hearing impairment as a consequence of carboplatin treatment15-19 even after long
term follow up20 (Table 1). Cumulative carboplatin doses, younger age at treatment
5
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 6
initiation, and radiation therapy were identified as potential risk factors for sustained
hearing loss post-treatment.13,14,21
Platinum-induced ototoxicity presents as bilateral high-frequency sensorineural
hearing loss, with increasing incidence and severity in response to cumulative
dosage.22,23 Hearing impairment has negative effects on quality of life, cognitive
development, and learning, particularly in paediatric patients24-26 who may have a
concurrent degree of visual impairment from retinoblastoma or its treatment. Inter-
individual variation in the development of carboplatin-induced hearing loss exists, but
missing are clear predictors of ototoxicity risk prior to treatment initiation.13
A number of studies27-30 have explored the relationship between polymorphisms in the
genes encoding thiopurine S-methyltransferase (TPMT), catechol-O-methyltransferase
(COMT), and ATP-binding cassette sub-family C member 3 (ABCC3) and hearing loss
in patients who received cisplatin treatment, with conflicting results and lack of
consensus.27-30 No reports on genetic associations with carboplatin was published but
as carboplatin and cisplatin are both platinum-based chemotherapeutic agents, These
associations require further evaluation before beingto determine its eligibility for being
considered for clinical implementation to stratify patients at risk of ototoxicity when
treated with platinum-based chemotherapeutic agents.
The aim of this study was to assess potential clinical and genetic predictive risk
factors for ototoxicity in children with retinoblastoma receiving carboplatin
chemotherapy. These factors may be useful to personalize therapy for retinoblastoma.
6
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 7
2 | METHODS
2.1 | Ethics
This study was scientifically reviewed and approved by the Research Ethics Board of
The Hospital for Sick Children. Consent for research use of banked DNA was provided
during sample acquisition for clinical retinoblastoma genetic testing. The study is in
accordance with the Declaration of Helsinki.
2.2 | Sample
Ninety-sevenSeventy one children with retinoblastomaretinoblastoma, who received
carboplatin chemotherapy between January 1991 and September 2012 at the Hospital
for Sick Children and had adequate pre- and post-chemotherapy audiograms, were
retrospectively studied for clinical features and variants in genes reported to be
associated with platinum chemotherapy ototoxicity, using archived DNA from blood after
clinical RB1 mutation studies. Adequate pre- and post-chemotherapy audiograms were
available on 71 children who were included in the full analysis.
2.3 | Clinical data
Data retrospectively collected included: age at diagnosis and at start of chemotherapy,
sex, laterality, RB1 gene test results (blood), eye stage(s) at diagnosis, chemotherapy
details (number of cycles, drugs used and cumulative carboplatin dose), number of
audiograms and further treatments (radiation or autologous bone marrow
transplantation).
7
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 8
2.4 | Ototoxicity assessment
Pure tone audiometry31 (ref) or audiogram was the most commonly used test for hearing
assessment in our cohort especially when assessing older children, it gives a visual
representation of behavioralbehavioural response to sounds of varying frequency (0 to 8
KHx on horizontal axis) and amplitude (decibels on vertical axis). It requires children
Child cooperation is required rendering soit sometimes difficult to perform in young
children. In this group, otpto-acoustic emissions32 (ref) waere s initially used where
different frequencies were utilized to stimulate the cochlea was stimulated at different
frequencies to determine the presence or absence of hearing impairment at these
frequencies, but withoutimpairment, without information regarding the decibel of hearing
loss. When the child becomes old enough pure tone audiometry was always performed.
Ototoxicity was scored in each audiogram by the audiologist in the form of a loss of
sound amplitude in decibels at specific frequency in kHz for each ear. The toxicity
grading for each child was then performed by three independent observers (authors CD,
SS and HD) using different grading systems based on the audiogram of the worse ear
at the last follow up.
The following toxicity grading systems were used: National Cancer Institute Common
Terminology Criteria for Adverse Events (NCI-CTCAE) version 333, Children's Cancer
Group (CCG), International Society of Paediatric Oncology (SIOP) Boston ototoxicity
scale34, and Brock35 and Chang36 systems to grade ototoxicity (Supplemental Table 21).
Patients with ototoxicity of grade 2 or higher by at least one classification system were
considered to have hearing loss, similar to inclusion criteria in previous studies.34-36
8
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 9
2.5 | Genetic analysis
Genotyped samples identified as heterozygous or homozygous for one or more of
five minor allele variants [TPMT (rs12201199 and rs1800460), COMT (rs4646316 and
rs9332377), and ABCC3 (rs1051640)] were used as positive controls to optimize real-
time PCR assays for detection of the five variants (Supplemental Tables 1-32-4). Two
sets of amplification reactions were designed for each variant using allele-specific
primers and real-time PCR was monitored by SYBR green dye. Differential amplification
efficiency for each allele was determined with 2 different sets of primers pairs based on
reference sequences37 (Supplemental material). Primers specific for the variant allele
were designed to have a melting temperature of 5C lower than primers for the wild-type
allele in order to enhance allelic discrimination (Supplemental Table 34).
2.6 | Statistical analysis
Ototoxicity (hearing loss) was treated as a dichotomized variable: grade 0 versus grade
2 (by at least one classification system). Classification schemes were deemed to be in
agreement if they produced equal grades for one or both ears, as applicable.
Two-tailed Fisher’s exact tests were used to analyse associations between hearing
loss and clinical characteristics including sex, diagnosis with unilateral or bilateral
retinoblastoma, age at diagnosis and age at treatment initiation. Association between
number of audiograms and cumulative carboplatin dose were evaluated using the
Mann-Whitney tests. Age at diagnosis and age at treatment initiation were also treated
as continuous variables, and the Mann-Whitney test was used to test for differences
between patients who developed ototoxicity and those did that not. Receiver operator
9
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 10
characteristic (ROC) analysis was used to determine how well age at treatment initiation
functions as a risk predictor for the development of ototoxicity post-treatment with
carboplatin in retinoblastoma patients. The true positive rate (sensitivity) and false positive
rate (100 – specificity) for different cut-off ages at treatment initiation to distinguish patients
with hearing loss (grade 0) versus no hearing loss (grade 2 by at least one classification
system) was used to determine area under the curve (AUC). The AUC is a measure of how
well age at treatment initiation can distinguish between children that develop hearing loss
and children with unaffected hearing after treatment with carboplatin for retinoblastoma,
3 | RESULTS
The clinical characteristics of the 71 included patients are summarized in Table 31. Sixty
one patients (85%) had bilateral retinoblastoma, 8 unilateral (11%), and 2 trilateral (3%).
The majority of patients were female (n = 40; 56%). The median age at diagnosis was 11
months (range, 6 days to 8.6 years). Sixty-nine patients (97%) were 3 years of age or
younger at diagnosis. The median age at treatment initiation was 12 months (range, 16
days to 8.8 years). Sixty-eight patients (96%) were 3 years of age or younger at the time of
receiving chemotherapy. Sixty-nine patients (97%) were 3 years of age or younger at
diagnosis. The median age at treatment initiation was 12 months (range, 16 days to 8.8
years). Sixty-eight patients (96%) were 3 years of age or younger at the time of receiving
chemotherapy. All patients received systemic chemotherapy with carboplatin, etoposide
or teniposide, and vincristine [vincristine and teniposide (n = 25; 35%), vincristine and
etoposide (n = 43; 60%), vincristine, etoposide, and teniposide (n = 2; 3%), and vincristine,
etoposide, teniposide, ifosfamide, and adriamycin (n = 1; 1%)], with concomitant
administration of high dose cyclosporine on the Toronto Protocol.38 Thirteen out of 71
10
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 11
patients (18%) received radiation and 4 patients (6%) were recipients of a bone marrow
transplant. The median cumulative dose of carboplatin for all 71 patients was 1,400 mg/m2
(range, 260 to 5,148 mg/m2 ).
3.1 | Hearing loss
Twenty-four of 71 patients (33.8%) developed mild to profounda degree of hearing loss
deafness ( grade 0 1 at least one classification system) at some time after treatment
initiation (median, 40.5 months; range, 3 months to 13 years). . Eighteen patients
(25.4%) had moderate to profound hearing loss ( grade 2 at least one classification
system) with a time to detection of ototoxicity between 2 and 75 months (median, 18.5)
from the start of treatment with carboplatin (Table 42) and were included in the
statistical analysis. Most patients with hearing loss experienced bilateral (15/18, 83.3%),
grade 1 or grade 2 ototoxicity. Three patients (16.7%) also experienced unilateral grade
1 or 2 ototoxicity. It was noted that all patients with grade 2 or higher hearing loss by at
least one classification system were assigned grade 1 or higher ototoxicity by all other
classification systems. Thus, no patient with grade 0 ototoxicity by any classification system
was included in the hearing loss group.
The median number of audiograms per patient was 7 (range, 1 to 25 evaluations)
including a baseline audiogram performed at the time of treatment initiation (Table 3).
On average, patients with hearing loss received more audiograms (median, 11.5
audiograms; range, 1 to 25) than those without hearing loss (median, 5.5 audiograms;
range, 1 to 19).
11
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 12
Hearing loss was observed to be grade 3 or higher (at least one classification
system) in 6 of 18 patients (33%). At time of submission of this paper, 3 patients wore
hearing aids.
3.2 | Agreement among classification systems
The greatest agreement was observed between the Brock and Chang classification
systems for ototoxicity (68 of 71 patients, 95.8%). The SIOP and Chang systems were
second highest in agreement (67 out of 71 patients, 94.4%) and the NCI-CTCAE
system was least often in agreement with the other classification systems
(Supplemental Table 21).
Figure 1 demonstrates a Kaplan-Meyer curve of time of occurrence of hearing loss in
patients with ototoxicity using different grading systems.
The NCI-CTCAE grade was lower than at least two other classification systems in 4
out of 18 patients (22.2%), the same as at least two other classification systems in 6 out
of 18 patients (33.3%), and higher than at least two other classification systems in 8 out
of 18 (44.4%) patients. Furthermore, 6 patients assigned grade 0 (no hearing loss) by
all other classification systems were assigned grade 1 ototoxicity by the NCI-CTCAE
classification system. The inclusion of these 6 patients in the group of patients with
hearing loss would result in an 8.8% increase (25% to 33.8%) in the incidence of
ototoxicity in this cohort
3.3 | Clinical and genetic risk factors for hearing loss
Potential clinical and genetic predictors risk factors of ototoxicity were assessed. In
univariate analyses, the only significant risk factors were age at diagnosis (P = 0.01)
12
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 13
and age at treatment initiation (p = 0.008) (Table 31). Eight out of 18 patients (45%) with
hearing loss were less than 6 months of age at treatment initiation.
ROC analysis identified age at treatment initiation of less than 4.25 months to have
the highest likelihood ratio for the development of ototoxicity post-treatment with
carboplatin in retinoblastoma patients. The area under the curve was calculated to be
0.7059 (p = 0.008). The sensitivity, specificity, and likelihood ratios for different cut-off
ages at treatment initiation to distinguish patients with hearing loss (grade 0) versus no
hearing loss (grade 2 by at least one classification system) are listed in
Supplementary Table 45. This finding was further assessed by constructing three
Kaplan-Meier curves for the development of ototoxicity after treatment with carboplatin
in children younger than 4.25 months (Figure 2a1a), 6 months as reported by Qaddoumi
et al.,24 (Figure 2b1b), and the median age at treatment initiation, 10 months (Figure
2c1c). The log-rank test was used to assess the difference between the two groups
based on the cut-off age at treatment initiation as specified. A significant difference was
noted when the cohort of patients that developed hearing loss was separated into
groups younger than 6 months (p = 0.0045) and 4.25 months (p = 0.0027), but not
younger than 10 months (p = 0.2103). Age at treatment initiation younger than 4.5
months and 6 months conferred odds ratios of 4.99 (95% CI, 1.3960 to 17.8002; p =
0.01) and 3.36 (95% CI, 1.0559 to 10.6922; p = 0.04), respectively.
There was no association between genetic variants in TPMT, COMT, ABCC3 and
ototoxicity (Supplementary Figures 1-2).
13
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 14
4 | DISCUSSION
In the current study, age at diagnosis and age at treatment initiation were identified
as the only statistically significant variables associated with hearing loss. Age at
treatment initiation was also a risk predictor of carboplatin-induced ototoxicity, where
younger patients were more likely to develop ototoxicity than older patients. This is in
accordance with previously published reports of clinical predictors of carboplatin-
induced hearing loss in paediatric patients.24,39 Baseline audiograms at the start of
treatment ensured that the ototoxicity identified in patients in our study occurred after
treatment with carboplatin was not pre-existing, unrelated to chemotherapy. As
Ototoxicity occurs in younger children, the careful titration of potential risks of different
treatment modalities40 including lower dosage chemotherapy (ref),41 Topotecan based
chemotherapy,42 cyclosporine free chemotherapy, (ref) periocular chemotherapy 43,44 or
intra-arterial chemotherapy45,46 should be performed. Furthermore, strict audiological
follow up during active treatment can easily identify early hearing loss and further
systemic chemotherapy can be re-evaluated based on tumour response and visual
potential.
Previous studies (Table 3) have reported predominantly bilateral (90%), grade 3 or
higher (90%) ototoxicity in retinoblastoma patients24, albeit with a higher cumulative
carboplatin dose (median, 3,576 mg/m2) received during chemotherapy than in our
study (median, 1,400 mg/m2). The incidence of ototoxicity following carboplatin-based
treatment in retinoblastoma patients is higher in our study (25%) than formerly
reported17-21,23-25 (0% – 16.7%, Table 13). Differences in methodology, definition of
ototoxicity, and heterogeneity between the cohorts have made it difficult to draw
14
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 15
comparisons. Chronic cyclosporine A has also been associated with the development of
hearing loss;39 our cohort all received short term (simultaneous with chemotherapy) high
dose Cyclosporine A to reduce the impact of multidrug resistance.8,9 In our cohort, the
chief difference in the chemotherapy protocols was standard or high carboplatin
dosage. Vincristine was a part of chemotherapy in concert with carboplatin and other
agents such as etoposide and teniposide. There was no significant difference in
vincristine dosage between the group of children that developed hearing loss (n = 18)
and those that retained normal hearing (n = 53). Chronic cyclosporine A has also been
associated with the development of hearing loss;47 our cohort all received short term
(simultaneous with chemotherapy) high dose Cyclosporine A to reduce the impact of
multidrug resistance.14,48 Utilization of multiple grading systems helped picking all
children with a degree of ototoxicity which was not utilized previously,
Our study is the third24,49 to employ multiple grading systems for ototoxicity, including
the most recently introduced Chang system of classification36. We demonstrate high
agreement (90.2 – 95.8%) among the SIOP, CCG, Brock, and Chang classification
systems. The NCI-CTCAE classification system has been reported to underestimate the
frequency of platinum therapy-induced ototoxicity.49 However, our data revealed that the
NCI-CTCAE classification grading alone resulted in overrepresentation of ototoxicity.
Genetic variants in TPMT and COMT were initially discovered as predictors of
cisplatin-induced ototoxicity in a candidate gene study of 166 paediatric patients with
different types of cancers27. Subsequent validation studies of the same variants in
TPMT and COMT revealed inconsistent results. Pussegoda et al.50 replicated the
association between TPMT variants and cisplatin-induced ototoxicity in a cohort of 155
15
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 16
patients. However, another study found no association between TPMT and COMT
variants reporting and ototoxicity in 213 paediatric patients with medulloblastoma.23
Ratain et al.,51 expressed concerns regarding the two published reports that
demonstrated an association between TPMT, COMT, and ABCC3 variation and
cisplatin-induced hearing loss27,52. These studies failed to support former provisional
patent applications, evinced discrepancies in the data, and may have overestimated the
significance of associations due to inadequate correction for population variation in the
polymorphisms that were examined.
Finally, meta-analysis of previously published studies27-29, in addition to two
independent cohorts of 100 Dutch and 38 Spanish patients with osteosarcoma, yielded
no statistically significant association between genetic variants in TPMT and COMT and
cisplatin treatment-related ototoxicity30. Our results do not support the hypothesis that
genetic variation in TPMT, COMT, or ABCC3 is associated with carboplatin-induced
hearing loss in retinoblastoma.
There are several limitations of this study mainly being retrospective, single centre
and non-comparative. The sample size was small mainly due to the genetic portion of
the study depending on the availability of banked DNA. The statistical power of our
study was limited due to the cohort size and expected minor allele frequencies (range,
1.28% to 21.81%) of the genetic variants that were evaluated. Baseline audiograms
insured the intact hearing at carboplatin treatment initiation but the confounding ototoxic
effect of vincristine and Cyclosporine A can’t be neglected as a possible cause of our
higher ototoxicity levels.
16
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 17
Genotype-phenotype associations, particularly between genetic variants and drug
toxicities, are often confounded by non-genetic factors. Increased cisplatin dosage,6
younger age,53-55 cranial irradiation,54,56 and use of aminoglycosides55,57,58 and
vincristine55,59,60 have been reported to affect platinum-induced ototoxicity.
The genetic basis of carboplatin-induced ototoxicity has yet to be determined. One
study identified a missense mutation in eIF3, which encodes the largest subunit of
eukaryotic translation initiation factor 3 (EIF3) and plays a role in DNA repair, as a
potential biomarker for cisplatin- and carboplatin-related nephrotoxicity and ototoxicity in
lung cancer patients.61 Several genetic pathways regulate the uptake, transport, and
clearance of platinum.62 As a result, genetic risk factors for platinum-related hearing loss
might well include more than one gene. In a comprehensive review of platinum-induced
ototoxicity in paediatric patients, Brock et al.,30 suggest the use of novel methodologies
for a “polygenic approach” to identify genetic determinants of hearing loss.34 Large-scale
genome approaches such as next generation sequencing (NGS) might elucidate
genetic risk factors for treatment-related ototoxicity. The use of NGS for the
identification of variants associated with ototoxicity after treatment with cisplatin or
carboplatin is limited. A recent genome-wide association study of 238 paediatric
patients with brain tumours discovered that variant rs1872328 in the enzyme-coding
ACYP2 gene was overrepresented in children who developed ototoxicity after treatment
with cisplatin.63 The authors replicated their findings in a cohort of 68 children treated
with cisplatin using targeted gene resequencing.
In summary, we show that treatment with carboplatin is accompanied by a significant
risk of bilateral, irreversible hearing loss, particularly in younger children. Additional
17
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
D’ Silva et al. Carboplatin ototoxicity in retinoblastoma patients 18
pharmacogenetic studies are required to ascertain genetic determinants of treatment-
related ototoxicity in retinoblastoma.
CONFLICT OF INTERESTS
No financial conflicting relationship exists for any author. BLG is unpaid medical director
for Impact Genetics Inc.
ACKNOWLEDGEMENTS
The authors would like to acknowledge Rob Laister and Mike Jain (Department of
Medical Oncology and Hematology, Princess Margaret Cancer Centre, Toronto,
Canada) who assisted with the molecular and statistical analysis during this study.
REFERENCES
18
352
353
354
355
356
357
358
359
360
361
362
D’Silva et al. Carboplatin ototoxicity in retinoblastoma patients 19
Figure Legends
FIGURE 1: Kaplan-Meier plot of time to development of ototoxicity (months) in
retinoblastoma patients treated with carboplatin. The majority of hearing loss (grade 0 or
higher) was detected between 3 and 30 months from treatment initiation with
carboplatin-based chemotherapy (n = 23). Each curve represents ototoxicity definition
by different classification systems: NCI CTCAE ≥ 0 (orange), Brock ≥ 0 (red), SIOP ≥ 0
(green), CCG ≥ 0 (violet), and Chang ≥ 0 (blue).
FIGURE 2: Kaplan-Meier (KM) plot of time to development of ototoxicity (months) in
retinoblastoma patients treated with carboplatin separated by age. KM curves were
generated for carboplatin-treated retinoblastoma patients and separated by age at
treatment initiation. (A) Younger than 4.25 months (blue) and older than 4.25 months
(red). (B) Younger than 6 months (blue) and older than 6 months (red). (C) Younger
than 10 months (blue) and older than 10 months (red). P-values are indicated in
boldface below each figure legend.
19
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
D’Silva et al. Carboplatin ototoxicity in retinoblastoma patients 20
Table Legends
Table 1: Summary of clinical characteristics and univariate analysis of the Cohort
(n=71)Published reports of carboplatin-induced ototoxicity in retinoblastoma
Table 2: Clinical characteristics of patients with moderate to severe hearing loss (n =
18).Grades of Hearing loss (ototoxicity) using different systems showing the degree of
agreement on ototoxicity definition in our cohort between them
Table 3: Summary of clinical characteristics and univariate analysis.
Table 4: Clinical characteristics of patients with moderate to severe hearing loss (n =
18).Published reports of carboplatin-induced ototoxicity in retinoblastoma
Supplemental material Legends
Detailed methods and technique of genetic testing are described in Supplemental
material and Supplemental Figures.
20
378
379
380
381
382
383
384
385
386
387
388
389
390
391