Abstract 2934


The analysis of polymorphisms in drug metabolism pathways and in DNA repair genes could help to identify patients with possible different treatment response and outcome. Single nucleotide polymorphisms (SNPs) are the most frequent type of genomic polymorphisms and have been described in association with prevalence, response to treatment, progression-free and overall survival in different tumors, including multiple myeloma (MM). The aim of the present study was to examine 22 SNPs related to DNA repair and drug metabolism, and correlate our findings with response, toxicity and survival in patients with MM after autologous stem-cell transplantation (ASCT).

Patients and Methods:

One hundred and eighty seven patients with MM (103M/84F, median age 55 years) intensified with melphalan-based ASCT have been studied in one institution. The median follow-up was 4 years (range 4 months to 18 years). None patient was lost to follow-up. Genomic DNA was isolated from bone marrow slides using a commercial assay (Qiagen). SNPs were analyzed by TaqMan assay in an ABI Prism 7500 Sequence Detection system (Applied Biosystems). The genes and SNPs evaluated in genomic DNA by allelic PCR were ERCC2 (rs13181, rs238406), ERCC5 (rs1047768, rs17655), XPA (rs1800975), XPC (rs2228001), XRCC1 (rs25487), XRCC5 (rs1051685, rs1051677), XRCC4 (rs963248), ERCC1 (rs3212948, rs735482) and BRCA1 (rs16941, rs799917) for DNA repair systems; NAT2 (rs1799930), CYP2C8 (rs11572080, rs2275622, rs10509681), TYMS (rs2790), SULT1 (rs1402467) and GST1 (rs1695) for phase I and II drug metabolisms, and ABCB1 (rs1045642) for drug transportation. These genes were selected based on their potential impact on prognosis in solid tumors in previous reports.


In the overall population, median PFS was 2.7 years (CI 95% 2.2 to 3.3 years), with a median OS of 6 years (CI 95% 4.5 to 8 years). OS was significantly shorter in patients with SNPs in ERCC5 (rs1047768; p=0.021), XPA (rs1800975; p=0.032) and GSTP1 (rs1695; p=0.015) (Figure). There was also a trend for CYP2C (rs2275622; p=0.054) and TYMS (p=0.107). The significance of the SNP in ERCC5 was retained in the group treated with conventional chemotherapy at induction (p=0.034), but not in those who received novel drugs (bortezomib, thalidomide and lenalidomide). Patients with SNP in ERCC1 achieved a lower CR rate (22.2% vs. 37.8%; p=0.033), with no prognostic significance.

Polymorphism in GSTP1 was also associated with a shorter PFS (p=0.002), without differences in the complete remission (CR) rate. When only patients who received ASCT after first line treatment were considered, the effect over OS remained at significant level (p=0.039). Furthermore, the effect on PFS and OS was also significant in patients achieving CR after ASCT (p=0.03). NAT2 (rs1799930) and ERCC2 (rs238406) polymorphims were associated with clinically significant mucositis after conditioning, as well as TYMS (rs2790) with relevant gastrointestinal toxicity (p<0.02). No other associations with prognosis or toxicities were observed with the remaining SNPs.


SNPs in differences DNA repair systems, such as ERCC5 and XPA, were associated with longer OS in patients MM after ASCT. Since these polymorphisms were not related to a better response or longer PFS, it can be speculated that the more prolonged OS could be due to a potential higher efficacy of rescue therapy. A SNP in GSTP1 (Ile105Val), a phase 2 drug metabolism enzyme involved in the metabolism and detoxification of a range of chemotherapeutic agents including melphalan, was associated with a shorter PFS and OS, as reported in previous series. Our findings could be useful to identify patients with MM who are more likely to benefit from melphalan-based therapies.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.