Children's Hearing Loss From Cisplatin Chemotherapy Associated With Gene Variants

Jacquelyn K. Beals, PhD

November 20, 2009

November 20, 2009 — Scientists at the University of British Columbia, Vancouver, Canada, have identified genetic variants associated with ototoxicity that occur in more than half of children treated with cisplatin. The report, published online November 8 in Nature Genetics, raises the possibilities that individuals at higher genetic risk for ototoxicity could receive lower cisplatin doses, alternative drug treatments, or be chosen for otoprotectant studies.

Cisplatin is a highly effective platinum-based chemotherapeutic agent whose anticancer action is the result of platinum complexes that bind and cause DNA cross-linking, eventually resulting in apoptosis. Adverse effects include nephrotoxicity, neurotoxicity, and ototoxicity. Hearing loss is reported in 10% to 25% of adults who receive cisplatin and in 41% to 61% of children. Hearing loss that predates language acquisition is particularly serious, and cisplatin therapy is often reduced or discontinued when ototoxicity occurs in young children.

Participants in this study were recruited by the Canadian Pharmacogenomics Network for Drug Safety, a consortium that studies adverse drug reactions in children. The discovery group consisted of 54 children from pediatric oncology units who had received cisplatin therapy. Among them, 33 had "serious cisplatin-induced toxicity," as noted in moderate to severe hearing deficit (loss of 25 dB or more). The others experienced no significant hearing loss.

Ototoxicity May Be More Common in Boys

A replication study of 112 additional children treated with cisplatin at pediatric oncology units throughout Canada found 73 with moderate to severe hearing loss. In the replication group, boys were significantly more likely to demonstrate ototoxicity (67% boys vs 50% girls; P = .042). Interestingly, ototoxicity was far less frequent in children with germ-cell tumors (6.6%) than with other pediatric cancers (26.8%; P = .0006).

"We are unsure of the cause of this [sex effect] and only observed this difference in the second cohort of patients," said first author Colin J. D. Ross, PhD, research associate, Department of Medical Genetics, University of British Columbia, Centre for Molecular Medicine and Therapeutics, and the Child and Family Research Institute, Children's and Women's Health Research Centre of British Columbia, Vancouver, via email to Medscape Pathology & Lab Medicine.

"As our work in this area continues, we will be further examining this gender difference," Dr. Ross said. The lower ototoxicity in children with germ-cell tumors is also of interest. "This may be related to different cisplatin administration protocols for patients with this tumor type," said Dr. Ross, "and we are following up on this finding now."

David R. Freyer, DO, MS, pediatric oncologist, director of the LIFE Survivorship and Transition Program, Children's Center for Cancer and Blood Diseases, Children's Hospital Los Angeles, and Department of Pediatrics, Keck School of Medicine, University of Southern California–Los Angeles, observed by telephone to Medscape Pathology & Lab Medicine, "I'm not sure if that was a reflection of some other covariants...the dose of the cisplatin, gender. Those germ-cell patients actually tend to be older, interestingly enough," he said. Dr. Freyer was not involved in the study.

"To my knowledge, there are no data that adequately explain why young children are affected more than older children are," Dr. Freyer continued. "Studies have shown that the breakpoint seems to be around 4 or 5 years of age. If you're younger than that when you get treated...there's more of an ototoxic effect than there is at an older age," he said. Hearing loss occurs in older children and teenagers but is more frequent in young children.

TPMT and COMT Associated With Ototoxicity

Nearly 2000 single nucleotide polymorphisms (SNPs) in participants' DNA were analyzed to identify variants of 220 genes that influence the "absorption, distribution, metabolism, and elimination" of drugs and drug metabolites, and any association with cisplatin ototoxicity. Variants of genes TPMT and COMT were highly associated with cisplatin-induced hearing loss.

A TPMT variant was found in 27.3% of children with cisplatin ototoxicity in the discovery cohort and 21.9% of those in the replication group; the same SNP occurred in none of the discovery control patients and in only 2.4% of replication cohort control patients (P = .00022; after Bonferroni correction, P = .032). A COMT variant was present in 33.3% and 29.2% of children with cisplatin ototoxicity in the discovery and replication groups, respectively; the same SNP occurred in 50.0% and 58.3% of children in the respective control cohorts (P = .00055; but after Bonferroni correction, P = .076).

DNA sequencing of these genes demonstrated 2 additional variants in TPMT that impair TPMT enzyme activity, and 1 additional variant associated with cisplatin ototoxicity in COMT (P = .000182; after Bonferroni correction, P = .0261). Cisplatin-associated hearing loss was more severe (P = .0000027) and had earlier onset (P = .00009) in children with more than 1 risk allele.

"We are enrolling additional patients in our studies to address this question [of allele effects], because of the rarity of some of the risk alleles," said Dr. Ross. "Only 3 patients were homozygous for TPMT risk variants, all...developed serious hearing loss. Similarly, only 5 patients were homozygous for the COMT risk variant, all...developed serious hearing loss," he noted.

In addition, the first TPMT risk allele mentioned occurred in 5% of people of European heritage, 1.7% of Asians, and 51.7% of Africans — the latter is consistent with the more common cisplatin-associated nephrotoxicity in blacks; data for ototoxicity in African populations are not known.

"Incredibly Important" Study

"This just incredibly important, and the data are very compelling — preliminary, but compelling," said Dr. Freyer. He mentioned a collaboration planned between the Vancouver-based group and the Children's Oncology Group study, which will validate the results in a larger patient population.

"Some of those patients who are being studied for the gene mutation are also going to be treated with this preventative medication, called sodium thiosulfate," Dr. Freyer noted. If a preventative effect is confirmed across the population, it will be interesting to see whether children with these gene variants are preferentially protected, raising the possibility of targeted interventions in the future.

"That can be the direction that this is going," predicted Dr. Freyer. "Identify who these patients are at risk, right up front, and then target interventions toward that group, those most at risk."

Dr. Ross reported that the study was funded as part of the peer-reviewed Genome Canada Applied Health Research Program; pharmaceutical industry partners (which include Pfizer, Eli Lilly, Merck Frosst, and Janssen-Ortho) had no formal or informal role in this program of research. Dr. Freyer has disclosed no relevant financial relationships.

Nat Genet. Published online November 8, 2009.