As I described in my prior post, the marker ERCC1 (excision-repair cross complementing group 1) is a prognostic variable that is associated with a more favorable survival in patients who aren't treated with chemo after surgery for early stage NSCLC. But this marker also appears to be predictive of resistance to cisplatin and a worse survival in patients treated with platinum-based chemo after surgery. In later stages of lung cancer, where platinum-based chemo is a common backbone of treatment, there is some work that suggests that ERCC1 expression is associated with worse results, although this line of research is still relatively early in development.

One of the early important papers that supported this conclusion was from Lord and colleagues, a collaborative effort among researchers in California and Spain, in which ERCC1 messenger RNA levels were evaluated in 56 patients treated with cisplatin/gemcitable (abstract here). The results demonstrated that high ERCC1 levels were associated with worse outcomes:

(Click on image to enlarge)

Again, this is unlike the finding in resected early stage NSCLC not treated with chemo.

This work was the basis for a trial called the Genotypic (molecular marker-based) International Lung Trial, or GILT, which randomized 2/3 of the approximately 300 enrolled advanced NSCLC patients to a customized treatment approach of either standard chemo with cisplatin/taxotere or gemcitabine/taxotere depending on whether ERCC1 levels predicted sensitivity or resistance, respectively, to cisplatin. This customized group is compared to the other 1/3 of patients who all received standard cisplatin/taxotere without any customization:

Early results indivated that the customized treatment arm was associated with a significantly better result (abstract here), and final results are anticipated to be published soon in the Journal of Clinical Oncology. I'll summarize the findings when this is published.

Several other trials presented at the World Conference on Lung Cancer just a couple of months ago in Korea also supported these results. A Korean study of 71 patients with stage IIIA N2 NSCLC who were treated with cisplatin-based chemoradation followed by surgery showed that those patients with low ERCC1 expression was associated with a significantly longer median survival than was seen among patients with high ERCC1 expression (65.1 vs. 20.5 months) and that ERCC1 was among the most predictive variables of better or worse survival (abstract here). Another trial in advanced NSCLC showed that high ERCC1 levels are associated with low survival particularly among adenocarcinomas (abstract here), but because this was retrospective and treatment details weren't available and was not uniform, interpretation of results in terms of response to chemo really can't be made. Finally, a couple of abstracts (here and here) also indicated that high ERCC1 expression is associated with worse outcomes among patients with SCLC, a setting in which platinum-based chemo is the standard.

I've oversimplified here, since there are studies that don't support a clear association of ERCC1 and outcome after treatment for advanced NSCLC (abstract here), and some of the studies noted above report on other potentially important markers as well, and in some cases the association of outcome and ERCC1 status wasn't astounding. Importantly, this is a setting where changing treatment based on ERCC1 would represent deviating from appropriate standards of care based on a non-standardized, still unvalidated test. Because of this, I haven't used this test yet, even though I suspect that with further study it may lead us to eventually alter our treatment plans, leading us to favor a non-platinum combination instead of standard platinum doublet therapy for patients with tumors that express high levels of ERCC1. But in the meantime, we await the results of the GILT trial and a good deal of ongoing work on ERCC1 and the ability of this marker to predict resistance to cisplatin-based chemotherapy.