Association of candidate pharmacogenetic markers with platinum-induced ototoxicity: PanCareLIFE dataset

Genetic association studies suggest a genetic predisposition for cisplatin-induced ototoxicity. Among other candidate genes, thiopurine methyltransferase (TPMT) is considered a critical gene for susceptibility to cisplatin-induced hearing loss in a pharmacogenetic guideline. The PanCareLIFE cross-sectional cohort study evaluated the genetic associations in a large pan-European population and assessed the diagnostic accuracy of the genetic markers. 1,112 pediatric cancer survivors who had provided biomaterial for genotyping were screened for participation in the pharmacogenetic association study. 900 participants qualified for inclusion. Based on the assessment of original audiograms, patients were assigned to three phenotype categories: no, minor, and clinically relevant hearing loss. Fourteen variants in eleven candidate genes (ABCC3, OTOS, TPMT, SLC22A2, NFE2L2, SLC16A5, LRP2, GSTP1, SOD2, WFS1, and ACYP2) were genotyped. The genotype and phenotype data represent a resource for conducting meta-analyses to derive a more precise pooled estimate of the effects of genes on the risk of hearing loss due to platinum treatment.


a b s t r a c t
Genetic association studies suggest a genetic predisposition for cisplatin-induced ototoxicity. Among other candidate genes, thiopurine methyltransferase ( TPMT ) is considered a critical gene for susceptibility to cisplatin-induced hearing loss in a pharmacogenetic guideline. The PanCareLIFE crosssectional cohort study evaluated the genetic associations in a large pan-European population and assessed the diagnostic accuracy of the genetic markers. 1,112 pediatric cancer survivors who had provided biomaterial for genotyping were screened for participation in the pharmacogenetic association study. 900 participants qualified for inclusion. Based on the assessment of original audiograms, patients were assigned to three phenotype categories: no, minor, and clinically relevant hearing loss. Fourteen variants in eleven candidate genes ( ABCC3, OTOS, TPMT, SLC22A2, NFE2L2, SLC16A5, LRP2, GSTP1, SOD2, WFS1, and ACYP2 ) were genotyped. The genotype and phenotype data represent a resource for conducting metaanalyses to derive a more precise pooled estimate of the effects of genes on the risk of hearing loss due to platinum treatment.
© 2020 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license.

Value of the Data
• This database describes genotypes of 14 candidate SNPs for cisplatin-induced ototoxicity in a large Pan-European cohort of pediatric cancer survivors treated with platinum. • Epidemiologists interested in the frequency of platinum-induced ototoxicity as well as developers of long-term follow-up guidelines for survivors of childhood, adolescent, and young adult cancers may benefit from these data. • The genotype and phenotype data represent a resource for conducting meta-analyses to derive a more precise pooled estimate of the effects of genes on the risk of hearing loss due to platinum treatment. Table 1 summarize the genotype and allele frequencies of the study population, stratified according to the hearing loss phenotype.

Data Description
The supplementary Table 1 shows demographic and clinical variables at a patient-level of the total cohort (n = 900). Variables are described as follows:  ID -Unique identification number assigned to each patient who was included in the analyses. RX -Cranial radiation (0) or no cranial radiation (1) SEX -Male (1) or female (2) PHENO -audiological phenotype: no hearing loss (0), minor hearing loss (1), clinically relevant hearing loss (2). Patients were assigned to the respective audiological phenotype based on the post-treatment audiograms of the patients, which were graded according to the Münster Classification. A detailed description of the phenotyping method is given below. AGE -age at start of platinum treatment: < = 5 years (1) Table 2 shows the genotype data at a patient-level of the total cohort (n = 900). Variables are described as follows:

Study design and participants
Background and methods of the European multicenter PanCareLIFE study have been described previously [1][2][3] . Patients were enrolled after approval was obtained from local review boards and written informed consent was obtained from patients, parents or legal guardians. Participants were enrolled both retrospectively and prospectively (i.e., chemotherapy was started and finished during the 5-year term of PanCareLIFE). Eligibility criteria were: 1) age at diagnosis < 19 years, 2) treatment with cisplatin, carboplatin or both, 3) at least one pure tone audiometry within 5 years after the end of chemotherapy. Exclusion criteria were: 1) non-consent and 2) hearing loss before the start of platinum treatment. Patients of this larger ototoxicity cohort participated in the pharmacogenetic study if there was additional consent for the genetic analyses and biomaterial was provided.
Laboratory assistants were blinded to the audiological phenotype of the patients. Multiple positive and negative controls and replicate samples were included in the genotyping assays and plates. No genotype discordance of replicate samples was observed. Ten samples were finally excluded due to genotype call rate per sample < 100%.

Audiological classification and phenotyping
All audiograms were independently rated by two reviewers for hearing loss according to the Münster classification [ 4 , 5 ]. Audiograms had to meet the following minimum requirements: frequencies include at least 2 or 3 kHz, 4 kHz, and 6 or 8 kHz (air-conduction), demonstrate no conductive hearing loss, absence of significant test artifacts (e.g., atypical air-bone configuration).
Thereafter, two pediatric audiologists independently assessed the kinetic course of hearing loss for each patient. The minimum data requirement for phenotype assessment included the availability of a normal pre-treatment audiogram or a normal audiogram before the third platinum cycle and at least one post-treatment audiogram within 15 months after the last chemotherapy cycle. Sound field audiometry was also accepted if ear-specific pure-tone audiometry was subsequently performed. Three phenotype groups were defined as follows: no hearing loss, minor hearing loss, and clinically-relevant hearing loss at the end of treatment. Patients were assigned to the no hearing loss group if post-treatment audiograms were exclusively Münster class 0. Patients were also assigned to the group without hearing loss if post-treatment audiograms were almost exclusively graded as Münster class 0, no audiogram was classified as Münster > 1, and the Münster class 1 audiogram was followed by a Münster class 0 audiogram. Patients were assigned to the clinically-relevant hearing loss group if follow-up audiograms indicated hearing loss of at least Münster class 2b. All other patients were classified as part of the minor hearing loss group. Inter-rater agreement was > 95%. After completion, all cases that had been phenotyped differently by the two pediatric audiologists were discussed between them and an agreement was made.