Introduction:

Chemotherapy induced peripheral neuropathy (CIPN) is a recognized clinical outcome of vinca alkaloid based therapy that frequently results in dose reduction or therapy discontinuation, but the determinants of interpatient variability remain unpredictable.

Previous pharmacogenetics studies have demonstrated that vincristine is metabolized through the CYP3A system with the CYP3A5*3 variant (GG allele at rs776746) having the least efficacy in metabolizing vincristine to its inactive metabolite M1 and conferring higher risk of neurotoxicity. Similarly, a single nucleotide polymorphism (SNP) in the promoter of the CEP72 gene (TT allele at rs924607) has been described as being significantly associated with increased risk and severity of vincristine-associated peripheral neuropathy in the pediatric population. Our recent studies have further studied the incidence of this risk allele and its correspondence with CIPN in the adult population treated with vincristine. This study aimed to investigate if an additive SNP analysis, both for CYP3A5*3 and CEP72, would determine increased predictability of CIPN in patients treated with vinca alkaloids.

Methods:

A retrospective case-control study of patients was completed in patients that were exposed to vinca alkaloids (vincristine, vinorelbine, and vinblastine) at Geisinger Medical Center from 2007-2015. Available samples of these patients were identified in the Geisinger MyCode bio banking project and the SNP of interest was genotyped using the TaqMan® SNP Genotyping Assay, with genotypic discrimination performed using SDS software (Applied Biosystems). Cases of CIPN were confirmed with retrospective chart review. Statistical analysis was conducted using SAS (SAS Institute Inc.) software, version 9.4. All genotyping and chart review was conducted by personnel blinded to sample identity.

Results:

A total of 121 patients had genotype mapping completed, of which 60 patients received vincristine, 33 received vinblastine, and 28 received vinorelbine. CIPN was detected in 35.5% of the study population. The CEP72 risk allele (TT) was detected in 14.9% (n=18) of the cohort and the CYP3A5*3 risk allele (GG) was detected in 89.2% (n=108) of the cohort, with similar incidence in each drug class. The median age of this population was 61 years of age with 52.9% (n=64) being males.

Multivariate logistic regression analysis, controlling for age and sex, did not demonstrate a significant model for the recessive genetic model of both SNPs with all vinca alkaloids (p=0.33), vincristine alone (p=0.26), vinorelbine alone (p=0.98), or vinblastine alone (p=0.97). The incidence of the inactive G/G allele for CYP3A5*3 was seen in 87.5% (n=21) of the vincristine cohort with neuropathy,

The incidence of neuropathy was higher in patients with the TT genotype (75%) in patients receiving vincristine therapy than with the CT or CC genotype (35.7%). As consistent with prior findings, a similar dominant effect was not detected in vinorelbine or vinblastine groups. In the vincristine population (n=60), diseases that were treated included diffuse large B-cell lymphoma (n=29), acute lymphoblastic leukemia (n=6), oligoastrocytoma (n=5), and Ewing's sarcoma (n=2). No disease specific variation in neuropathy incidence and association with TT genotype was detected. Patients received a median three doses of vincristine at 2 mg each (total 6 mg) before detection of CIPN. Four patients (6.67%) of the patients required dose adjustment of chemotherapy, including dose reduction (n=2) or early discontinuation (n=2).

Conclusions:

This study demonstrates that the CEP72 genotype confers predictive modeling for patients receiving vincristine-based therapy and that including CYP3A5*3 analysis in an additive model is not associated with an increased predictive model for CIPN in patients receiving vinca alkaloid based therapy. In our population, the high incidence of the inactive G/G allele of CYP3A5*3 made it difficult to discern a significant effect in patients that experienced neuropathy. Larger studies are needed to include a more diverse population to confirm this predictive association and to understand if this can help guide precision-based chemotherapy administration and favorably impact disease prognosis.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution