e18115 Background: New therapeutic approaches are being developed based on the findings that several genetic abnormalities underlying NSCLC could influence chemosensitivity. We assessed whether polymorphisms in genes of NER pathway, such as ERCC5, ERCC6, MMS19L, CCNH, XPC, RRM1, can affect the tolerability of platinum-based chemotherapy in advanced NSCLC pts. Methods: Eligible pts had histological or cytological confirmed stage IV or IIIB (with malignant pleural effusion) chemo-naïve NSCLC, ECOG PS of 0 to 2 and adequate organ function. Pts received vinorelbine 25 mg/m2 or gemcitabine 1000 mg/m2 on day 1 and 8, or docetaxel 75 mg/m2 or taxel 175 mg/m2 on day 1 plus cisplatin 75 mg/m2 or carboplatin AUC=5 on day 1, every 21 days per cycle, for a maximum of six cycles, unless disease progression or unacceptable toxicity occurred. Hematology and biochemistry analyses were repeated every cycle or before chemotherapy for d8 treatment. Toxicity was classified according to CTC AE 3.0. Venous blood was collected and genomic DNA was isolated. Polymorphisms were assessed. SNPs in ERCC6 (A/G 1413, A/G 1213 and A/G 1097), XPC (T/C 500 and A/C 940), ERCC5 (G/C 1104), RRM1 (−37C/A), CCNH (T/C 270) and MMS19L (G811A) were assessed. Results: Totally 388 pts were included prospectively from Oct 2005 to Mar 2009. Median age was 61 years; 68.8% were male, 36.2% were ECOG PS 1, 33.2% had stage IIIB. The median follow-up time was 18 months (range 8–66 months). Hematological toxicity was the most common AE. SNPs of MMS19L and ERCC6 can predict the AE of chemotherapy in NSCLC,leucopenia (P=0.020), jaundice (P=0.039) by MMS19L. In terms of grade 3-4 AE, SNPs of MMS19L and RRM1 -37C/A have much more predictive effect. SNPs of MMS19L were good at total AE (P=0.042), thrombopenia (P=0.035), SNPs of RRM1 -37C/A were for vomiting (P=0.046) and jaundice (P=0.014). Conclusions: Polymorphisms in ERCC6, MMS19L, RRM1, can predict the tolerability of platinum-based chemotherapy in advanced NSCLC patients. Polymorphisms in other genes of NER pathway, such as ERCC5, CCNH, XPC, have no predictive effect.
Read full abstract