· web viewretrospective analysis of clinical and genetic associations for carboplatin related...

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

mailto:[email protected]


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


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


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


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


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


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


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


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


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


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


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


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


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

Sameh Soliman, 08/08/17,

How would the results affect your clinical practice?


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


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


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


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

· web viewretrospective analysis of clinical and genetic associations for carboplatin related...

Documents