ERCC1-Tailored Chemotherapy in Lung Cancer: The First Prospective Randomized Trial

The cancer death toll is one of the top health problems facing the industrialized world today. Lung cancer is the leading cause of cancer death in men and also increasingly in women. Novel therapeutic strategies need to be developed and existing therapies optimized to increase the survival rate of lung cancer patients. Improving systemic chemotherapy will have an impact on virtually 90% of all lung cancer patients, result in the cure of more patients with resected non–small-cell lung cancer (NSCLC), and improve the survival of patients with locally advanced/metastatic lesions. In this regard, rational treatment decision making based on an analysis of biomarkers of response and resistance to cytotoxic drugs seems to be a promising approach. Platinum has long been the mainstay of chemotherapy for lung cancer. Platinum cytotoxicity results from the disruption of the double-stranded DNA molecule in cells, mainly through the formation of intrastrand adducts. Nucleotide excision repair (NER) is the primary DNA repair mechanism that removes platinum-DNA adducts from genomic DNA. Excision repair cross-complementing 1 (ERCC1) is a critical gene on the NER pathway. A growing list of reports links cisplatin, carboplatin, and oxaliplatin resistance to ERCC1 mRNA levels in tumors. This relationship has been suggested for patients with gastric, bladder, ovarian, colorectal, and non–small-cell lung cancers. We recently showed that immunohistochemically evaluated ERCC1 levels are also predictive for the survival benefit afforded by adjuvant cisplatin-based chemotherapy in patients with totally resected stage I to IIIA NSCLC. In the treated group, ERCC1-negative tumors derived a substantial benefit from adjuvant cisplatin-based chemotherapy, whereas their ERCC1-positive counterparts did not. Interestingly, in patients randomly assigned to the observation arm, the subgroup with ERCC1-positive tumors had a better survival compared with those with ERCC1-negative tumors. This paradoxical status of ERCC1 (a good prognostic marker in untreated resected NSCLC patients and a poor predictor of efficient adjuvant chemotherapy) was recently confirmed by Zheng et al. The study by Cobo et al in this issue of the Journal of Clinical Oncology evaluated ERCC1 mRNA expression prospectively in an attempt to predict response to cisplatin-based or cisplatin-free regimens in stage IIIB or IV in NSCLC patients. In the control arm (no customization), patients received the well-established docetaxel/cisplatin combination. In the genotypic arm, patients were allocated to docetaxel/cisplatin or docetaxel/gemcitabine regimens respectively, according to whether low or high ERCC1 mRNA levels were detected. The study met its primary end point: response rates were significantly higher in the genotypic arm (50.3%) when compared with the control arm (39.3%). However, this difference was no longer significant after a strict intent-to-treat analysis. Most importantly however, the clinical relevance remains limited, given that there was no difference between the two arms in either progression-free survival or overall survival. How can one explain the negative results of such an innovative pioneering trial of customized chemotherapy? Technical issues clearly are crucial in the setting of biomarkerbased therapies. The study by Cobo et al is based on a series of assumptions. Evaluation of mRNA gene expression by reverse transcriptase polymerase chain reaction may be assessed safely in a convenient material: paraffin-embedded tissue. Laser capture tumor microdissection guarantees optimal tumor enrichment of the analyzed specimen. This quantitative technique produces a ratio of ERCC1 mRNA expression with housekeeping genes, thus simplifying allocation by predefining a threshold. In the trial under consideration, however, approximately 18% of the patients allocated to the genotypic arm were inassessable for ERCC1 due to insufficient tumor material. This percentage, along with patients excluded for other reasons (such as brain metastasis, protocol violations, and so on), led the authors to amend the protocol and increase recruitment by nearly 30%. The 18% dropout rate emphasizes the poor feasibility of this technique when insufficient tumor tissue is available. ERCC1 assessment was performed by Response Genetics (Los Angeles, CA), proprietor of the mRNA extraction procedure. The exact threshold defining high ERCC1 expression levels is unknown, but 43% of the patients were allocated to the high expression arm (high genotypic arm). It is therefore surprising that ranges for ERCC1 expression valve overlap between the high and low genotypic groups. Overall, major improvements of quantitative mRNA assessment are still required in the prospective setting. This technique should be compared and put into perspective with immunohistochemistry, a semiquantitative method that can be performed with a minimum amount of tissue and that has been validated for ERCC1 evaluation by two independent groups in the NSCLC setting. The design of the study by Cobo et al also merits discussion. The underlying hypothesis is that ERCC1 customization should improve clinical outcome (response rate and survival). This should be achieved by offering cisplatin-based therapy to ERCC1-negative JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 25 NUMBER 19 JULY 1 2007