404MO - Clinical relevance of MIR27A rs895819 polymorphism and its interaction with DPYD variants for predicting grade 4-5 fluoropyrimidine (FP) toxicity (tox) in the FUSAFE individual patient data meta-analysis (IPD-MA)

Background

Dihydropyrimidine dehydrogenase (DPD) deficiency is the main cause of severe FP-tox. DPD expression is regulated at post-transcriptional level by miR-27a. MIR27A rs895819A>G has been associated with significant increased grade (G) ≥3 FP-tox only in patients (pts) carrying a deleterious DPYD variant (*2A/D949V/*13/HapB3). French GPCO-UNICANCER and RNPGx groups initiated the FUSAFE IPD-MA to develop predictive models of FP-tox, allowing to test whether MIR27A rs895819 improves the predictive value of DPYD variants.

Methods

Eligibility criteria included unbiased recruitment of Caucasian pts without FP dose adjustment based on DPD status. Main endpoint was hematological or digestive G4-5 FP-tox at 12 weeks. A clinical logistic regression model including age, sex, body mass index, FP administration (bolus±continuous vs continuous/p.o.) and associated anticancer drug (AAD) was first developed. Combined DPYD variants *2A/D949V/*13 with or without HapB3 (at least one mutated allele) and MIR27A rs895819 (dominant coding) were further included in this model, along with DPYD x MIR27A interaction.

Results

Full clinical data, MIR27A, and DPYD variants were assessable in 2724 pts (5 studies): 71% colorectal cancer, 41% metastatic, 95% FP-naïve, 31% 5FU, 55% AAD. MIR27A rs895819G was carried by 53.4% of pts (9.7% homozygotes) and 1.5% of pts (N=41) carried at least one *2A/D949V/*13 DPYD variant. G4-5 FP-tox was 4.0%. Adjusted Odds Ratio were 6.9 [95%CI 3.0-16.0] for DPYD combined variants *2A/D949V/*13, and 1.3 [95%CI 0.9-1.9] for MIR27A rs895819, with no DPYD x MIR27A statistical interaction (p=0.99). Results on MIR27A and its interaction with DPYD variants were similar when considering G3-4-5 tox at cycle 1, or HapB3.

Conclusions

This IPD-MA on MIR27A and DPYD variants, including the largest series of pts published so far, do not confirm the previously published relevance of MIR27A rs895819 for toxicity risk stratification in patients carrying DPYD risk variants. The ongoing FUSAFE-2 project including sequencing of DPYD and MIR genes may provide new insights in FP pharmacogenetics.

Clinical trial identification

CRD 42017058148.

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

French Cancer Institute & French Health Ministry (PHRC-K 14-193 \"FUSAFE\") and French Ligue Contre le Cancer.

Disclosure

M-C. Etienne-Grimaldi: Honoraria (self), Travel/Accommodation/Expenses: Amgen. V. Boige: Honoraria (self), Advisory/Consultancy: Bayer; Honoraria (self), Advisory/Consultancy, Research grant/Funding (institution): Merck Serono; Honoraria (self): Amgen; Advisory/Consultancy, Travel/Accommodation/Expenses: Sanofi; Travel/Accommodation/Expenses: Daiichi Sankyo; Honoraria (self): Roche; Honoraria (self): MSD; Honoraria (self), Advisory/Consultancy: Ipsen; Advisory/Consultancy: Prestizia; Honoraria (self), Advisory/Consultancy: Eisai. D. Meulendijks: Shareholder/Stockholder/Stock options, Full/Part-time employment: AstraZeneca. C. Palles: Advisory/Consultancy: Oxford Cancer Biomarkers. U. Zanger: Licensing/Royalties: Robert Bosch GmbH. J. Taieb: Advisory/Consultancy: Amgen; Honoraria (self): Lily; Advisory/Consultancy: Sanofi; Honoraria (self): Roche; Advisory/Consultancy: Merk; Honoraria (self), Advisory/Consultancy: MSD; Honoraria (self): Celgene; Advisory/Consultancy: Servier; Honoraria (self): Sirtex; Advisory/Consultancy: Pierre Fabre. E. Gross: Licensing/Royalties, Method for predicting treatment response in breast cancer: Patent pending. All other authors have declared no conflicts of interest.